You're forgetting Matthew 7:6 "Do not give that which is holy to dogs, and do not throw your pearls before pigs, for they will trample them under their feet, and turn and tear you to pieces."
I like talking with solicitors and missionaries, but there's basically zero chance I will ever attend their church or convert to their religion or whatever. I think I'll try referencing the phrase next time someone is being overly insistent on giving me a pamphlet or bible (which will immediately end up in the recycle bin).
Re: 1 - I've always had a morbid interest in the record holders of human suffering, and while I used to think the worst serial killers were likely to be those who killed the most, I now believe certain gang members (particularly from Latin America) or death squad perpetrators (some of whom were interviewed at length in the documentary "The Act of Killing", but one can also think about militia-type murderers during the Rwandan genocide for example) could dwarf the numbers of even the most prolific "lone wolf" type serial killers.
The key is that Soviet executions involved a lot of people doing repetitive assigned roles to facilitate many killings per hour and this guy's role was "fire the gun"
if it was state sanctioned then it doesn't meet Scott's definition of "illegal". Legal killings are always going to vastly outnumber illegal killings. Still interesting, but in a different category.
War crimes are illegal even if a state sanctions them, and the Katyn Forest massacre certainly qualifies. That's an awful lot of murders, done up close and personal one bullet at a time. And Blokhin wasn't just "the guy assigned to fire the gun", he was the guy assigned to organize the effort and he *chose* to hog all the gun-firing glory for himself.
Wishing? Everyone agreed that it was a war crime, and it was even included in the Nuremberg trials. The Soviets just didn't manage to fabricate enough evidence to get the Germans convicted for it.
I think Scott would say that "state sanctioned" killing would be outside of his criteria for that link. Even agreeing that it was illegal, it was orchestrated as part of a state's official orders. If you include that, then the numbers can skyrocket and aren't comparing the same things.
Auschwitz would also count, and whoever was pushing the buttons for the gas chambers, for instance. Though Blokhin might still be the worse individual!
If we're counting wartime, there's Simo Hayha at 505, but sniping in a declared war is definitely legal rather than illegal.
I think the highest kill counts in the sense of "pushed the button/pulled the trigger themselves" are Thomas Ferebee, the bombardier on Enola Gay who dropped Little Boy on Hiroshima and Kermit Beahan, the bombardier on Bockscar who dropped Fat Man on Nagasaki.
Well, most Western governments have rules restricting the killing of wild animals, but also send government agents to kill specific animals that are known to have killed humans.
Right but they also kill animals who haven’t hurt anyone, simply for being “too close” to human settlements. This frequently happens to bears and wolves in North America. In Australia, saltwater crocodiles are closely monitored and “harvested” when they risk encroaching on human settlements.
Pretty sure it was against the Japanese law of the time to drop a bomb on a Japanese city. (Yes, this is a silly argument, but it's also a bit silly to refer to the law when talking about the actions of wild animals going against the human law.)
That’s as far as I read but yes, please. More of that. Great idea. I have no idea why elevated rail didn’t work, I assume it was economic, but hopefully this will.
Elevated rail works fine - it's just expensive, takes up a lot of space, and very noisy. But it's still a large part of what is getting built.
Gondolas are a nice idea, but they haven't yet caught on anywhere other than mountainsides (including Bogota, where they have become significant urban transportation, because the city is on a mountainside).
It remains to be seen whether Sugar Land will actually start building any.
Gondolas might make a lot of sense in certain places across water. Maybe Hong Kong? Loads of people catch the ferry across the 700m span between Kowloon and Hong Kong Island, but a gondola would be faster (turn up and go, no waiting!)
Looking at the map for Sugar Land I'm not sure it's an ideal case. The problem with Sugar Land is that the "downtown" is basically four malls at the intersection of two freeways. While I can see there's some value in being able to quickly get between these four malls I'm not sure if the gondola is the most sensible way (right now there's not even a pedestrian bridge). I can see some tourism value, but honestly the view of mall carparks and freeways isn't going to be that great.
I just remembered the other place I've seen a public transport gondola in a flat location: Singapore.
Well, I haven’t been on it in years so I don’t know. But it was definitely a commuter used way of getting to Roosevelt island because it’s only the F train that stops there.
It's about 2/3 resident/commuter traffic. More touristy than, say, the bus but nothing like the San Francisco cable car or other novelty transit.
The F train stops at Roosevelt Island and is convenient for some destinations but the tram is better if you're heading somewhere on the east side (can transfer easily to the 4/5/6).
Roosevelt Island is also the most popular place to live for UN staff/diplomats, and the UN is a fairly easy walk south if you take the tram.
> Loads of people catch the ferry across the 700m span between Kowloon and Hong Kong Island, but a gondola would be faster (turn up and go, no waiting!)
While the original use case was ski resorts & mountainous cities, gondolas work equally well in flat urban environments. They retain their inherent advantages of minimal infrastructure & low build cost.
The problem with Sugar Land is that it is the 6-th fastest growing US city and literally grew so quickly that there's not enough land left to fit in transit. It's not intended to be a tourist ride. The city transportation manager describes it this way: "“[When] people see it in action, I think it will be transformative to how people get around. It's public transit that people would actually want to use. A third of our population doesn't even have a driver’s license,"
And Swyft Cities uses on-demand autonomous technologies, unlike conventional gondolas. And it is orders of magnitude less expensive than any other transportation form, even a pedestrian bridge.
What was your experience like with the gondola in Singapore?
I mean Sugar Land isn't really a city anyway, it's an outer suburb of Houston. There's no point in trying to build a public transport system just for Sugar Land, you need one to cover all of Houston.
Agree with everything you're saying. A different blog today said "Houston is so big there are, well, cities inside of it." While it needs a metro-wide system, the Houston-The Woodlands-Sugar Land Metropolitan Area is a mind-boggling 10,000 square miles, bigger than Rhode Island, New Hampshire & Connecticut combined.
Even within Sugar Land, there is a need for district-level transit that would connect several clusters of retail, office & residential that are cut off from each other, esp. by I-69 & Hwy 69. So these neighborhoods are close to each other but difficult to go between.
Modern elevated rail is not particularly noisy. NIMBYs hear about elevated rail and think of scenes in movies establishing our character is poor and their life sucks by having the noisy train clank through right outside their window and wake them up... but modern elevated rail sounds like this: https://youtu.be/lthbSSC00ek?si=o-vx6s46LnZGpeId
The key difference vs. a traditional gondola is that the cables don't move and are arranged as a network, and the vehicles can seamlessly switch cables to shortest-path through the network.
So like a gondola in form, but more like aerial railcar in function.
Not just Bogota but also La Paz (I've been there, and it works). The one US city that pops in my mind is Pittsburgh - surely that's a plausible candidate for gondolas, or alternatively a Wuppertal-style system?
Amazing what La Paz has done using gondolas as their primary transit system. Mexico City, Quito, Medellin, Santo Domingo & several other LatAm cities use gondolas as part of their multimodal transit networks
I like the Chicago L a lot, but, yeah, kind of disruptive and expensive.
Doesn't Portland have a little public transport gondola going up a hill? It looks like public transport users can get a pass for $440/year https://www.gobytram.com
Portland's tram connects the original OHSU campus up in the hills with the expansion campus at the south waterfront. The hospital gives free tickets to patients who have to go between campuses.
As with all public transport the key questions are:
Capacity (measured in passengers, per direction, per hour). Gondolas tend to be ~ a tenth the capacity of a subway line.
Speed: not actually a massive issue in most cases - subways rarely average much about 20-30 mph, note that's *average,* including wait time at stations and acceleration/deceleration times, maximum speed is usually 50-60 mph.
Independence of traffic - gondolas are fine for this: streetcars and buses can be badly affected (though they can also have dedicated lanes / dedicated rights of way and be fine too)
Independence of weather - gondolas usually have an upper limit for the wind speed
And then a bunch of minor factors like ease of access, accessibility for disabled passengers, etc
Note that the main complaints about elevated rail (as distinct from other modes) are noise and also privacy (if the rail passes in front of people's windows, passengers can see into the upper stories of buildings either side of the road). Gondolas will be quieter, but I suspect the privacy issues are going to be at least as bad and possibly worse (gondolas are more open and can see into more stories).
I live in Canada. There's basically no place here where gondolas wouldn't be affected by wind/snow/ice. Yes, I know they're like ski lifts, but ski lifts are designed to only run in the winter, and their capacity is much lower.
Of course, there's a craze for light rail here, which combines the problems of buses and subways. They run on the surface, so they're subject to weather. They run down the center of the road, so reducing its capacity (this may be a plus for the planners). And they're really expensive and take a long time to build.
I don’t know about how well light rail copes with deep snow (heavy rail is generally fine because the rails are so high that you just need a snowplow) but reducing the capacity of a road for cars generally doesn’t make congestion worse as long as there is an alternative (and the light rail is exactly that alternative) - enough drivers will switch to alternatives that the congestion will usually end up less bad.
And light rail really shouldn’t be that expensive. Should be about a fifth to a third of the price of a subway (depends on whether you grade separate the junctions with crossing roads).
Snow isn't a problem; ice is. It's a problem for both light and heavy rail.
If there were a dense enough mass transit network that you could actually get to the light rail line, and get somewhere after you get off, then it would be useful. As it is though, you're likely walking. As opposed to automobile, where you park nearby.
I suppose you could outlaw parking, which would be one way to force people onto mass transit.
No, the light rail route that's being built outside my window will not be adopted by anyone who's not already using the buses currently running on that route. Since the trains will be larger than the buses but run less often, their utilization will be about the same - crowded during rush hours, empty otherwise.
Light rail may be cheaper, but it is less durable than subway due to more weathering. There's another light rail on the other side of the metropolis, and they're having to rebuild it after about 30 years. Meanwhile the subway serving the metropolis is still running fine.
Capacity - subways & light rail draw most riders from only 1/2 mile radius of stations. Our goal is to feed more riders from surrounding neighborhoods to help rail transit work more effectively
Speed - we actually have a huge advantage over light rail & buses, which only average 8-18 mph because of frequent stops. Even Uber/Waymo is slowed by traffic lights & congestion. Our gondola-like vehicles are autonomous and navigate independently across fixed cables. All trips are nonstop from origin-to-destination with no intermediate stops. So total trip times are faster than almost anything short of an express subway.
Independence - we love BRT & streetcars, but we help move traffic almost completely off of the ground plane, reducing congestion, which could also help BRT and streetcars move more freely
Weather - we use dampening technologies that expands our window for wind speed
Accessibility - our vehicles are completely stationery with level boarding, unlike conventional gondolas. No steps or gaps making it easy for elderly, disabled, wheelchairs, strollers, etc. Fully ADA-compliant.
Privacy -- ability for riders to see into buildings is definitely a concern for all elevated transit and we are working with cities and developers to avoid or mitigate concerns.
On capacity, your vehicle spacing is limited by the braking distance, and a gondola will swing about if they are being braked hard, so there are physical limits on how rapidly they can slow down. This is why most public transit systems try to use very-high-capacity vehicles; even a simple two-car light-rail will have a capacity of 150-200; a long subway train may have a capacity into four figures.
If you’re operating nonstop origin-to-destination then you’re going to have to use small vehicles, as only a relatively small number of passengers will share an o/d pair.
I’ll be genuinely worried that fairly small vehicles with an adequate safety spacing is going to mean you’ll need multiple cables to get a useful system capacity. Perhaps that’s the plan - have one cable per lane on the road below - but if not, then you’re looking at just not being able to move that many people (and - note - the total number of people you can move is also the maximum total revenue you can bring in, so if you can’t get many people, then you need to charge pretty high fares). Given that most of your cost is going to be in the support structures, running more than two cables between each support might be affordable if you can achieve some reasonable usage goals.
I note from your website that you do mention “pull-off stations” which makes a lot of sense - the only way that nonstop separate routing can work is if the stations are not on the main cableway (else the stopped gondola at a station stops all those following it).
Speed is great, though if passengers have to wait for a gondola going to their specific destination, then overall trip times may not be as fantastic as you hope; you’re going to need a lot of gondolas. How long a wait will there be if ten people arrive together with ten different destinations? And if there’s a lot of people waiting, they can’t even line up as they need to group together by destination. You’ll need to put some serious thought into designing the stations.
Big fan of the accessibility approach; you’re going to be looking at needing elevators at most of your stations, which may end up biting you a bit cost-wise (public outdoor elevators need a lot of maintenance and cleaning).
On the weather and privacy fronts, I hope it works out well for you: I suspect that you’ll need to select cities to target based in part on how windy they are (ie Chicago, perhaps not the best choice?). This is true of most forms of transit - urban geography is a major impact on which cities should pick what modes (e.g. Miami, with its high water table, would not be well-advised to build a subway). I do hope you have an approach to taking down the cables and securing everything when there’s a hurricane or tornado warning.
Given how automated your vehicles are, I suspect that being able to blackout the windows on one side of the vehicle may be a useful solution to the privacy problems - you can automate it so passengers can see forward and back and could look across the road, but not into the windows of nearby buildings.
Capacity - we're definitely not trying to compete with subways or even light rail, which are both great for moving high volumes of people. Our goal is to provide medium capacity that can move people around a neighborhood, and as part of that, feed more people into the neighborhood's subway & LRT stations.
Spacing - yes, braking ability is critical for vehicle spacing, esp. for safety considerations. Our key is vehicle control software that can dynamically adjust spacing between the vehicle ahead and the trailing vehicle. Speed, span length, turns, etc., all factor in. The good news is, 1-5 passenger vehicles running continuously can achieve surprisingly high capacity.
Nonstop trips - sorry, I may not have explained that well. Because each vehicle moves & navigates independently, when you request a ride on the app or at a kiosk, you select a destination. You board your designated vehicle, and the trip is then nonstop to your destination.
Stations - the offline pullover stations give us a lot of flexibility in station sizing. They can be single vehicle stations, or have multiple boarding berths for high traffic locations. And a station can be expanded as demand grows.
Stations can be ground level (think bus stop size) or elevated as needed. For elevated stations, cost, maintenance & cleaning of elevators is definitely a factor.
Stations can also be built alongside or even inside the upper stories of buildings. I live in Minneapolis which has an extensive downtown pedestrian skyway network, and I could envision our stations directly connected to the skyways.
Weather - wind is certainly a consideration, not just for operations but also passenger comfort if wind sway becomes excessive. We have dampening technologies that expand both the operational and passenger comfort envelopes, but certainly there eventually becomes an upper limit. Infrastructure (support towers, cables) is not an issue -- traditional gondolas cut their teeth in the Alps dealing with hurricane-force winds. We would likely berth & shelter vehicles during extreme conditions.
Privacy - Smart side windows like some airliners have that can dynamically darken or become opaque are definitely something we're evaluating for privacy.
This tech does seem really cool. However, as someone who lives in a city with good public transport - the occasions where I wish for a local neighbourhood transit option are when I'm trying to transport something - taking pets to the vet; shopping; DIY materials, taking stuff to the recycling centre, etc. That's fine if you can find places where these vehicles can 'land' but I think in many cities that will be a problem,and you'd usually have to resort to staircases,which makes this more of a faff. Indeed, even finding places to build a staircase could be an issue.
All good points. We're not perfect for every location, as every location has different transit needs & challenges. That said, one advantage we have is that we're very flexible in station configuration. Stations can be either ground-level or elevated. Ground-level stations can be as small as a corner bus stop. Our goal is to be accessible as possible for any location.
Also, with level boarding it should be much easier to bring a cart, small trolley, stroller, etc. onboard without to navigate it through aisle, around other people, etc.
Your project seems awesome, thanks for taking the time to answer nerdy questions here
If I'm understanding it right, the innovation is that you're doing away with tractor cables? I imagine one of the main benefits is that you have to splice any loops, and also you can probably have a more meandering path with less worries about tension.
Having spent many hours designing movable platforms for Poma, I'm puzzled about the assertion that traditional gondolas are not wheelchair accessible. I feel like the floor is perfectly level with the platform, and the bumper is resting against it so the gap is minimal anyway.
Could you explain what differentiates your product here?
Yes, unlike traditional gondolas of a single moving cable loop with vehicles affixed to it, we have a network of fixed cables and rail intersections. Each vehicle moves and navigates independently across the cable-rail network. And that then uses a completely different model for cable splicing and tensioning.
Sorry, I did not mean to imply that traditional gondolas are not wheelchair accessible. Most modern gondolas of course are ADA-complaint with level boarding and no or minimal gaps. Some OG ski lift use cases don't fit that model, but those are becoming increasingly rare. What I meant to say is that our vehicles are completely stationery while in-station, making boarding even easier for wheelchair, elderly, etc.
Probably a dumb question, but why is elevated rail a thing instead of elevated bus right of way? If it was the drivers, it seems like that should be updated since driving a bus on an elevated rail can definitely be automated at this point. It also seems like bus infrastructure would be much cheaper and quieter.
1. Buses are much noisier than rail (rubber on ashphalt is a lot louder than steel on steel as long as you don't have over-tight corners; even if you do, you can put automated lubrication systems on the corners so the wheels don't screech). ICE engines also put out noise, but I'd assume a new system would be electric for either bus or train.
2. Buses on guideways cause much more wear on the guideway than rail. You don't notice this on a road because they don't run exactly down the same route, but guided busways wear hard. They're usually solid concrete (ie no ashphalt) and they still need much more maintenance than steel rails.
Now, you don't have to use a guideway; you can just have a regular road surface and then your buses aren't positioned as precisely (which spreads the wear over the surface). Brisbane has a bus system exactly like this and it works well.
I was ready to like this, and the tech is cool, but don't see this as remotely viable without either extremely high pricing or high subsidies. Mainly due to the point-to-point part.
As far as I can tell the intended niche this is supposed to be filling is the same as that of a taxi or autonomous taxi. It is:
1) On demand (like a taxi)
2) Semi-point-to-point, still have to walk to and from a station (so somewhat worse than a taxi)
3) New infrastructure dependent
4) Elevated and "skips the traffic"
5) Autonomous (like a waymo)
This makes sense! The whole problem with mass transit networks in places like Sugar Land is that the size and densities just don't make a lot of alternatives particularly viable.
I initially thought that this would be a carriage sized gondola system (maybe with like SF trams level of capacity) which probably makes even less than this. Think about pricing and capacity on this taxi/car killer for a second.
You get to a Downtown station after a day of work. You're on your own- your colleagues all drive. There is a bit of a line. Someone at the front of the line shouts "is anyone else going to New Street station?". No one replies. You shuffle forward. Everyone individually, or maybe in their little group, gets in their own gondola and whizzes off at *up to* 30mph. Of course it's rush hour, so in fact the network is full and therefore a lot slower, and constantly stop starting due to people getting off. As you get to the front of the line a light flashes. "GONDOLA SYSTEM FULLY OCCUPIED- MEGA SURGE PRICING ACTIVE". You should have just got a waymo.
I'm being a little facetious, and all this stuff can (and does happen!) with cars. But it is absolutely clear on pricing that this would be competing with Waymo and Uber. You remove some of the human cost, sure, but the added in infrastructure prices (unless you get some nice city government to subsidise this boondoggle via the taxpayer) means I think it is unlikely to net out cheaper. Mass transit is something completely different - it serves dense, busy routes, with huge demand but sacrifices some flexibility.
It's not like the taxi line at LaGuardia ("Anyone else going to Midtown??") It's all on-demand. Here's how it would actually play out:
1. You live, work or are near a station
2. You request a ride on the app (similar to Uber or Waymo)
3. At the station, you board your assigned vehicle directly, which is waiting for you
4. You board and immediately depart
5. Your trip is private and nonstop direct to your destination
6. Because stations are all offline, people getting off and on does not affect mainline speeds, which remain constant.
Infrastructure costs will be a fraction of any other transit mode. And capacities are surprisingly high and more flexible depending on demand and location, unlike fixed rail infrastructure.
Also, and importantly, we do not compete with mass transit like subways. As you mention, they're fantastic for moving large numbers of people. We add the flexibility that they lack, helping to make mass transit more effective by feeding more people into stations very cost-effectively.
Thanks for the reply- is the idea then that there would be a much larger number of gondolas than potential riders, or at least ~around max demand? It seems cool, but I think from what you've said you'd agree that it is effectively a taxi or even cycling competitor. Assuming you'll be dynamically pricing (just makes sense with something like this) I do somewhat struggle to see how this scales up beyond a relatively low ridership, as the cables will be somewhat constraining. I guess the networked AI will help.
Re: #1: Isn't it likely that the person with the most directly killing is some kind of state-backed executioner? Charles-Henri Sanson performed almost 3,000 executions in-person as the executioner of the French king and later the French republic, and I think the guillotine is individual enough to still count, as opposed to say the guy who simply dropped the atom bomb onto Hiroshima. Another contender might be someone doing shootings for the Soviets, though I don't know how distributed the shootings were there.
Gustave of the Nile has not murdered any humans -- murder is defined in nearly every jurisdiction as an act of *homicide*. If we were to tally Gustave's homicide victims, we'd only look at the number of crocodiles that he's killed.
(Given Gustave's massive size, age, and visible battle scars, it seems highly probable that he has committed homicide; male crocs are fiercely territorial, and often fight other crocs to defend their turf. Crocodiles have also been observed engaging in acts of cannibalism.)
Since asexual reproduction makes the concept of "individual" a bit dubious, perhaps you could consider all clonal copies of the original /Yersinia pestis/ to be all instances of the same individual...
https://en.wikipedia.org/wiki/Vasily_Blokhin -- Reportedly killed at least 7,000 people personally in the Katyn Forrest massacres. Timothy Snyder talks about him in Bloodlands, apparently they just brought people in one at a time and he shot them with a revolver, for months and months. Appalling.
Oddly enough, Sanson did not enjoy being an executioner, and wrote papers on how to make sure that executions would be painless. I seem to recall he was from a family of executioners - he wanted to be a doctor, but had to drop out of college and inherit the family mantle when his father became crippled. Makes executioners sound a lot like haberdashers.
Gondolas: yes, that has been done at several places. I've used the system in La Paz, which has a particularly good reputation - indeed it seems to be a perfect match for the topography. La Paz is basically in the bottom of a bowl, many people live in the cold plain above the bowl (El Alto), and getting from one to the other because of work caused terrible jams and took forever. The gondola-based system seems to work nicely and gives good views of the city to boot.
Vance is definitely misusing the idea of "ordo amoris" (order of love). Augustine uses it to mean the "rightful" order of love, where what and how one loves is loved in the proper manner. Vance uses it to mean "hierarchical" order of love, where the people we are to love are ranked in order along the lines of Maslow's hierarchy of needs. Obviously there is some overlap, but overall Vance isn't using it quite correctly.
It really seems like the idea Vance meant to use was "subsidiarity", the idea that things should be handled at the lowest practical level and not by a centralized authority if possible.
The same way Catholics mean by "rightly ordered" love to mean heterosexual sex open to children. The idea is that sex is good in itself but to remain good it must be open to children or it becomes "disordered". (I wish there was a less controversial example, but this is the way most people have heard this usage of "order".). CS Lewis defines ordo amoris as "the ordinate condition of the affections in which every object is accorded that kind of degree of love which is appropriate to it."
Obviously you can make the argument that the appropriate degree of love for illegal immigrants is zero (as Vance does). Thomas Aquinas said that generally one's first responsibility was usually to those closest to us, but depending on need one might put the needs of a stranger in desperate need of help over the less urgent needs of one's own father. To put it another way, Augustine is referring to the idea of love properly oriented towards God vs Vance's idea of a ranking of who we love based on proximity. I hope that makes sense.
Augustine: “Further, all men are to be loved equally. But since you cannot do good to all, you are to pay special regard to those who, by the accidents of time, or place, or circumstance, are brought into closer connection with you.”
Vance: "You love your family, and then you love your neighbor, and then you love your community, and then you love your fellow citizens in your own country. And then after that, you can focus and prioritize the rest of the world.”
I'm not seeing a real difference here either, I have to say...
...except that—if your quote/paraphrase is accurate—Vance specifies what this means in more detail / provides more fine distinctions. Surely, though, the upshot is pretty much the same—i.e., your family is closer & more dear to you than are your neighbors or co-workers (who are, themselves, still closer to you than are random foreigners #49392–49395; etc.)—and, if we're using this principle at all, surely ranking love-duty by closeness /within/ the Augustinian categories ("those who ... are brought into closer connection with you" & those who... aren't), as well, is a fairly intuitive & justifiable extension.
Aquinas is saying there are practical reasons to favor those closer to you but not moral ones. Today the practical considerations are largely gone and in fact have reversed due to a dollar going further in poor countries, so Aquinas thought would lead to being an EA.
In Augustine's day, you couldn't really do much good for people hundreds or thousands of miles away from you. But that did not mean you should love them less or rank them as less deserving of help; if someone with malaria turned up on your doorstep of course you should help them as you would help a family member.
Those near to us are easier to help. And those near to us include the beggars in the streets, the sick and elderly and the strangers around us.
Is the difference that Augustine says you should in theory love everyone equally, but it's more practical to help those closer to you - but you're interpreting Vance as saying that even in principle, you should love those closer more? Or something else?
Yes, that is what I’m saying. Vance seems to think that our ENTIRE attention needs to be directed at those closest to us vs balancing need, proximity, responsibility, etc as Augustine (and Aquinas) says.
I don't know anything about US politics other than what I read here (and I must confess I do tend to skip or skim that stuff..) but if this Vance bloke is a politician that advocates for cancelling overseas aid programmes in order to make at-best-extremely-marginal domestic improvements at the cost of vast numbers of foreign lives, I do think that "you're interpreting Vance as saying.." is an overly-skeptical way of expressing it: if Vance is indeed one of the politicians that is cancelling the aid, I can't see many alternatives to Calvin's "interpretation".
Aquinas was writing at a time where you couldn’t help people further away even if you wanted to, and when the inequality between different societies was not so stark.
Applying his idea that you should love everyone equally but help the people you can help most easily to today’s circumstances means Effective Altruism at least in the old sense, because you can do a lot more good in much poorer countries where a dollar buys a lot more marginal utility.
Vance’s interpretation seems to go directly against the spirit of Aquinas’s point.
Yeah, strictly speaking Vance's definition is at odds with Augustine's, but I don't think Augustine's is internally consistent ("love everyone equally, just don't behave as if you did?")
Not to say that I especially agree with Vance's complaint here. (I mean, sure, concentric circles of moral concern are reasonable, but AFAICT the US government has *always* spent vastly more on domestic programs than foreign aid or even foreign wars. And physically importing illegal migrants is a fantastically inefficient/dangerous way to achieve humanitarian goals, so humanitarian concerns don't justify mass immigration.)
I don't see why "love everybody equally but focus on helping those whom you're in the best position to help" would be logically inconsistent? I love Marmite and Vegemite with perfect unbiased agape equality but Marmite is available at the shop on the corner whereas Vegemite is several hours' drive away, so it seems perfectly logical for me to mostly put Marmite on my toast?
>(I wish there was a less controversial example, but this is the way most people have heard this usage of "order".)
Wild speculation: Is there a culinary example? Maybe something about dining varying from sustenance to gluttony??? (I'm not religious so this is a _very_ ignorant guess.)
"Too much of a good thing". It's not wrong to love sports, but if you spend every free minute watching sports, going to games, talking about sports with your friends, and even go so far as to "sorry, can't come to our kid's school play because there's a match between the lowest ranking members of the lowest divisions I want to see", that's inordinate love.
You should love your mother, but you should also love your wife. And if you always take your mom's side, don't listen to what your wife says, always go "this is not how mom does it/cooks that meal/washes my shirts", and in short put more time and effort into being a son than a husband and father, that's inordinate love as well.
The one example most religious people would be familiar with is having idols instead of God as the rightful end of our love (where "idols" can be anything from literal idols to wanting more money/sex/status/sports merchandise/you name it).
As an aside, this is my own view and take on where Dante puts Bruno Latini in Hell; not for the sin of sodomy, but he is in the circle of the violent against God because he taught his pupils to pin their hopes on and put their trust in worldly fame, instead of (as their teacher) leading them to the greatest good, which is God:
Could Augustine’s point be roughly summarised as “loving people who are far away without loving those who are close first isn’t the right kind of love (disordered in some way)”?
I feel like it is more important to love fellow Americans (theoretically, of course since it is impossible to love a category) than it is to love a foreign communist who thinks he has the right to interfere in American politics. When protestant America wants the pope's opinion, as the old joke goes, we'll beat it out of him.
If a Catholic American politician makes a claim about Catholic theology but the Pope says that he was mistaken, I hardly think that counts as "interfering in American politics". If an American politician said that π = 3 and some foreign (presumably communist) mathematician said that actually π = 3.141592... would that be interfering in American politics?
(Also: I hesitate to help lower the useful:inflammatory ratio of the discussion to near-Twitter levels, but I find this an interesting enough idea that it's nevertheless worth mentioning:
There have been 250+ Popes, and the USA has been a country for maybe the last dozen or so of them. It's not-impossible that, should human civilisation last long enough, eventually the only evidence that the USA ever even existed might be records of the Pope's references to it in some dusty vault under the Vatican. Possibly recorded in some long-dead ancient language like Latin or English. Given ideological conflict between the USA and the Catholic Church, I know which side I'd bet on, in the long term!)
The Catholic Church is, AFAICT, the oldest continually existing organization on Earth, and even just counting continuous operation from the Vatican, they're a couple hundred years ahead of the US (which is not doing great ATM). So far, not having their own army has worked out fine for the Vatican.
I have asked two Catholic theologians and they thought Vance was right, Catholic theologians I talk to is not a random sample but I get the impression that the pope isn't seen as an uniquely powrful authority when not speaking ex cathedra.
reading about it, i think it may be closer to paul in 1 cor 7 about "it is good not to marry." he is using a negative sense.
the ideal for paul is not to marry and not to have sex as to serve the lord, but to avoid sexual immorality you should marry and in marriage only abstain from sex for brief periods.
the ideal is high but reality is not everyone can without falling into sin. paul would say the highest aim is to be "married to God." Being married is still a love, but not the highest love.
i remember reading that its actually very spiritually dangerous to be a monk for that reason. you have to be able to love responsibility or you become a nut and are worse off than if you aimed lower.
i don't think paul said that as a hierarchy of values though.
2 cor 8 is what i think you mean, and its more "you excel already in other aspects of the christian life, do not forget giving."
he also is pragmatic: their plenty will supply their need, and in turn when you need, they will supply you. the aim isn't to make you poor while others are relieved but to try to equal it out.
i mean i don't think it was held like the concept above was
Re: 10 - I wouldn't be surprised if Musk had multiple people run his account (after all he allegedly hired people to play videogames for him), which would partly explain why he's been able to post nearly incessantly over the past few weeks. The level of sleep deprivation suggested by the tweets/time graph would be difficult to sustain for more than a few weeks, right?
Re: 25 - YouTuber DIYPerks made an "artifical sun" in his home in a way that doesn't seem like it would need to cost $1000/object if it were made at scale and commercialized, and it has the benefit of looking like an actual window. The Brighter lamp is just that, a floor lamp, I'm not sure how convincing the effect would be - like sure, it'll be bright, but it'll still come from a distinct spot on the ceiling and get diffused from above, right? Hardly what sunshine indoors looks like in my experience
There is also that one weird gene expression that allows people to operate healthily on significantly less sleep than the norm. I wouldn't be surprised if its fairly common among uberachievers. Modafinil and Adderall only delay the paying of the debt.
I don't like the graph because it squeezes 365 days into less than 46 pixels horizontally. That's less than 1 pixel per week. If he stays up late on some nights and gets up early on others during a week, what does that look like?
(Tweeting just before bed and just after waking up is a problem but one that lots of people have.)
I'm still curious how he gets anything at all done with that much X engagement. Not just his own posts, but how much he likely reads from other people! I can't imagine getting anything at all done, let alone what appears to be a busy schedule even for high-achiever's standards.
Hiring people to send his tweets makes a lot more sense to me.
Re: 3 - I heard from someone who worked at the FCC at the time that the reason Net Neutrality ended when it did (and didn't have major consequences) was a result of the shifting corporate balance of power. Big Tech companies had fought and lobbied hard against repeal (partly, I suspect, astroturfing social media) because they saw the ISPs as peers, companies of reasonably similar size who were a real risk if they could take a bite out of tech's pie. By the time Net Neutrality was actually repealed, Big Tech had totally eclipsed ISPs in financial and political clout, and thus knew that ISPs wouldn't even attempt to fight them in the way Net Neutrality advocates feared. Big Tech was happy to shrug their shoulders and let Net Neutrality go simply because they no longer needed it, hence why nothing happened 2017-21.
Not really? You still can pirate pretty much every type of content with minimal hassle, something that governments, legacy media, big tech, small tech - everybody opposes. To the extent that people who don't want to use big tech have no good alternatives, it's because their interests are too niche, basically by definition.
"it’s probably worth it for middle-class people to spend $100 spamming a hundred companies with their applications, and any price high enough to discourage this would make it hard for poor people to apply at all."
But that's fine, though. If the cost of reviewing a crappy application that is an obvious poor fit is less than $1, then you'd be happy to get all the crappy applications people want to send out. Well-qualified poor people aren't hurt by this. In fact, it guarantees companies have the resources to give them a fair shot in reviewing their resume.
In the original, the money is donated to charity. Also, I don't think companies have a lot of slack in trading money for application-review time, since applications usually need to be reviewed by managers who are familiar with the company and its needs.
Currently most applications are reviewed by *software* - isn't this the original problem you described, since the reviewing software favours conventional, less-than-perfectly-honest applicants with the right keywords rather than necessarily good applicants? - which presumably produces some sort of list for managers to pick from.
If the cost of having a human or AI do this first-pass review, rather than a keyword filter program, were less than £1 per application I do think that charging £1 to apply would plausibly address the problem.
I mean, if you have an actually competent AI, none of this would even be necessary in the first place. The cost of having them go through applications is trivial.
Aren't all applications also written by AIs by now? The arms race is going to be legendary! This is exactly the kind of absurd future to be excited about.
Many school districts demand letters of reference, sometimes from a person's current(!) administrator, before even considering their application, electronically or otherwise.
I review hundreds of applications a year. Doing a first-glance approximation cost-effectively at $1/application would be fairly easy. You don't need to be particularly thorough, just separate the "obvious no" candidates from the pool and then pass them along.
You wouldn't want front line managers doing that, but you could easily afford to have additional recruiting staff sift through applications in that manner. It would give you approximately one minute per application, if you're paying an hourly rate of $30 and after loading benefits and additional costs of hiring the reviewer. You could get closer to two minutes per application if you're only paying $12-15/hour.
As it stands, that's a big expense without the $1 per application. Companies getting too many, especially obviously bad, applications need a cost-effective way to sort these applications otherwise, which I guess is where software comes in. My current organization gets around it by having a moderately difficult application process. If we were just going by applications on Indeed we would be overwhelmed sorting crappy drive-by applications from people who don't know or care what we even do.
It's not about paying for the cost of reviewing the application. It's a speed bump to make it harder for literal spammers to spam applications. Much like a captcha. Even my small company has to deal with this. A significant portion of our applications for engineering jobs are from services that file fake applications on behalf of foreign clients. Sometimes the people that comes to the interviews are clearly not the person who "applied". Even a nominal $1 charge would cut down on this only because they would have to provide a payment method. You could probably get the same effect without even charging and just requiring a credit card number.
Re 15. Although I don't believe omega-3 supplementation has any benefit in psychosis, I also don't think this new trial should shift your opinion much, given the total sample size was n=135 and the total number of transitions to psychosis was n=8.
The twin cities of El Alto/La Paz in Bolivia have an elaborate cable car system which is widely used for public transportation (https://en.wikipedia.org/wiki/Mi_Telef%C3%A9rico) -- using it when I visited in summer 2019 was kind of surreal, because you could look down and see the enormous class divides in the city (rich people live down in the valleys where there's literally more oxygen; the flat plain of El Alto is above 4 km in altitude!) while the walls of the cable cars had propaganda for the left-wing government in power at the time, which seemed grossly discordant to me. (But the free speech situation was fine -- our friends there felt totally free to shit-talk the government in those cable cars. This was a few months before the military coup which happened that winter).
Anyway, cable cars for public transportation, incredibly cool idea which has been implemented in some places!
Right, I visited pre- and post-cable-car, and there was a big difference - it's a beautiful system that works. And people felt free to shit-talk the government, which had overstayed its welcome in constitutionally dubious ways but seemed relatively benign compared to previous Bolivian governments (which had a knack for massacring their population, as the same friends and colleagues shit-talking the government also pointed out) and, well, um, let's not talk about what's going on elsewhere.
Agree, I first rode a gondola as part of a public transit system in Barcelona. Seamless transfer from the subway, and it was an unforgettable experience that made me an early believer in Swyft Cities well over a decade later.
Re: 22 - This seems like a classic toxoplasma of rage situation, although with substantially less distinct forms than those in the original post. It's a positive feedback loop, where each side defects harder and harder.
Re: 27 - Maybe some trans-related issues fall here? Although I'm not really sure (the main supporters of the statement "trans women are women" tend to be women, so this goes against that view).
There isn't a massive difference between men and women in terms of whether they see gender transition as morally acceptable. Men are less likely to see it as acceptable, but that's likely confounded by men just being more right-wing. Of course, the biggest predictor of transgender acceptance is just political alignment, to no one's surprise.
Re 22: emphatically yes. I think this loop has been running for at least 15 or 20 years in the U.S., probably quite a bit longer. I think changes in media technology and consumption patterns (especially but not exclusively social media) have supercharged it. I don't have much hope that the U.S. will be able to break out of it at this point, because all the obvious tools to do so have already been tightly captured by the loop.
I suspect the fight over “gender ideology in schools” is lower-male vs. upper-female. Education remains female-dominated.
But it might just be sampling bias. I’m exposed a lot more rants from blue-collar men. I had no luck finding polling on CW situations like Florida HB 1557 or the Loudon County policy.
For 1. The mongols used to masacre the populations of cities that didn't submit, and each soldier was given a quota of civilians to execute with their sabre. I expect most cities have larger populations than a mongol army had soldiers. An unusually bloodthirsty mogol that had been at a few sieges must have tallied up a lot of kills by hand.
>The Brighter lamp is 50,000 lumen ... My only concern is that the light costs about $1,000;
This is confusing to me... I already have a 21000 lumen bulb in my floor lamp, that I got on Amazon for $40.
Is there some reason that one 50k bulb would produce 'more light' in some relevant metric than just running 3 of those 21k bulbs?
Or is the price difference really just about the glare/temperature/etc concerns you mention? That seems like an extreme difference in price for just those concerns.
I also have a much cheaper bright bulb. I think it's a combination of glare/temperature/etc, form factor (eg a whole lamp is more expensive than a bulb), and probably pitching to a more premium market of well-off people who don't want to buy their own weird bulbs.
Do these lights come on slowly? How does the eye adjust? I can feel over powered by the relatively bright ceiling lights coming on when I’m relaxing in darkness or near darkness and somebody comes in and hits the switch, but those same lights are not obviously doing anything in bright sunshine. And we can’t stare at the sun, so how comfortable is staring at these lamps if possible.
In my experience, they come on fast, you go 'Aaaargh!' and look away for a few seconds, then you adjust and are fine.
But definitely you avoid staring at them, mine is in the corner mostly behind where I sit, and has a lamp shade diffusing the light and directing it towards the white ceiling.
So yeah it's possible these ones will be noticeably more functional in some ways. Still hard to justify the price difference, but maybe they start as a luxury and the price drops auick as you scale up manufacturing.
Is the point here brightness on dark days, because if I liked brightness at night I’d visit an airport. What I like at night is candles, or that equivalent of lighting. Grey days are depressing though.
The spectrum of the light (related to color temperature but much finer) is important. That costs more. A light that bright in an enclosed space sounds like something out of a black ops site.
Hey, creator of Brighter here. I find my eyes adjust quite quickly (2-5 seconds?), especially if some other lights are on. If you're wondering about dimmability, they're dimmable to about 5% (2500 lumens). You can control this through your smart home app (along with the color temperature.
Specs on amazon are basically the wild west. No one tests anything and almost everything is a lie. Flashlights are notorious for this, with random 6 letter brands competing to see who can come up with the highest number that someone will believe. So unless it's a quasi-reputable brand I would start off being skeptical of the claimed lumen level.
If it _is_ a reputable brand, then all those other factors like temperature etc. probably come into play.
This is the conundrum I've been facing. I would very much like to buy some very bright bulbs, but Amazon is simply no longer to be trusted for electronic items like that. Even if I were to pay extra for a reputable brand name, there's a decent chance I'd receive a Chinese knockoff that might, but probably won't, be as good as the genuine article.
So I stop by hardware stores every so often and take a look a their light bulb aisle. I expect that the supply chains for physical stores are a bit less infiltrated by IP-law-infrigining knockoffs, although perhaps my confidence is misplaced.
Same. I got a 30,000 lumen work light for ~$100. That said, the glare is pretty bad. I imagine someone handier / with more time on their hands than me could diffuse the light quite easily though.
The one other time I encountered this "lol just use more light, and it works" was on the old SSC so I'm surprised no one referenced it yet.
> Eliezer’s wife Brienne had Seasonal Affective Disorder. The consensus treatment for SAD is “light boxes”, very bright lamps that mimic sunshine and make winter feel more like summer. Brienne tried some of these and they didn’t work; her seasonal depression got so bad that she had to move to the Southern Hemisphere three months of every year just to stay functional. No doctor had any good ideas about what to do at this point. Eliezer did some digging, found that existing light boxes were still way less bright than the sun, and jury-rigged a much brighter version. This brighter light box cured Brienne’s depression when the conventional treatment had failed. Since Eliezer, a random layperson, was able to come up with a better SAD cure after a few minutes of thinking than the establishment was recommending to him, this seems kind of like the relevant research community leaving a $20 bill on the ground in Grand Central.
That may be true for the Amazon, in particular, but globally, the pattern is less clear. Wikipedia contradicts itself (https://en.wikipedia.org/wiki/List_of_countries_by_forest_area) starting by stating that forest area increased between 2010 and 2020, but then presenting a table with slightly different figures, showing a decrease.
According to either set of figures, forestry seems relatively stable, with forest coverage increasing over the period by 3.5% according to the first figures or decreasing by 1.2% according to the second figures.
Globally perhaps, but not all forest is the same. Rainforest like the amazon is in swift decline pretty much everywhere. Small comfort if Boreal forest is increasing.
The most commonly cited practical reason would be medicines usually. You never know what Gila Monster is out there waiting to give you the next Ozempic.
For me I'll just say killing the Amazon rainforest feels like a crime against the planet, and against wonder. Like blowing up Everest or killing the buffalo. I believe in the duty to preserve.
I've heard it's more complicated than that, in an interesting way. Venezuela, for example, is known for its tepui - their version of a mesa. These formations are high enough that their tops are effectively isolated from the surrounding area, and will evolve their own ecologies.
It gets even better. They frequently have natural sinkholes due to the properties of the rock forming them. Those sinkholes can be quite large, and sport micro-ecologies of their own, all isolated from each other.
I get the impression that if something were to wipe out the entire Amazon but somehow leave these tepui untouched, the total number of extant plant and animal species might only drop by a few percent. (I have no known reliable way of precisely measuring this, however - even biologists don't fully know what's in all those sinkholes.)
>The toxic bite of a Gila monster can kill a human, but a specific ingredient in the cocktail of the lizard's venom is the reason we have glucagon-like peptide (GLP-1) agonists like Ozempic and Wegovy.
Evidence over the decades suggests that the Amazon rainforest was itself a human construction for around 11,000 years. (Look up "terra preta".) A rough sketch of the narrative now is that there were as many as 5-8 million people living in what is now the rainforest in pre-Columbian times, perhaps fertilizing and nurturing thousands of square miles of gardens, which ran wild after their numbers declined.
If this is true, then destroying the rainforest might be a crime against pre-Columbian humans at worst.
Humans have lived in and altered the Amazon for about 11,000 years, but the forest itself is over 50 million years old, and at no point did they alter the entire thing. I don't think you could make any claim that the entire Amazon was a human construction at any point, or anything close to it.
Again, look up "terra preta" and related articles. I see no definitive evidence that the rainforest was 50 million years old, and strong evidence that it was densely populated for over 10 thousand years. Presumably, a significant fraction of the land is covered with artificial soil; although no one's been able to perform a comprehensive survey, regions previously believed to be virgin turn out to be growing out of terra preta.
Density and diversity. An acre of rainforest contains a lot more mass of plant life, and contains a much wider variety of both plant and animal species, than an acre of monoculture farmed trees that are be harvested and replanted every fifteen years. But they both count as "an acre of forests".
1) Higher rate of CO2 processing Vs boreal forest due to faster growth (although boreal can have a higher *stock* of stored CO2 per area unit, so it's complicated)
2) More importantly in my view but also more subjectively is the biodiversity in rainforests, which one can make utilisation arguments for (e.g. medicines) but which I think is intrinsically valuable. A universe with an Amazon rainforest is one I value more than a universe with dozens of barren planets or a world tiled solely in kudzu.
A square mile of rainforest holds an order of magnitude more animal and plant species than temperate ones, sometimes pushing towards two orders of magnitude. So, if we were trying to protect only biodiversity as such, we would focus only on rainforests and abandon the others almost entirely.
This is correct. Even if you limit yourself to just the Amazon rainforest, there's so many species that recording and classifying them all is hard. Most of the stuff you hear about "species going extinct before science could record their existence" will be about rainforest species.
A while back I believe the numbers floating around were 25% of Earth's terrestrial species living in rainforests, even though rainforests only cover ~5% of the planet. Now various websites are saying 50% of species, but I'm not sure where they're getting that number from.
The big thing is that we want a diversified portfolio of ecosystems. Preserving one sort of ecosystem has some comparability with preserving others, but it’s better to have some preservation of many ecosystems than to have a lot of preservation of just one ecosystem.
>But where is the symmetrical working-class women vs. upper-middle-class men gender war?
Given that you just used the example of upper-middle-class men not wanting to cheat with the Nanny and considering the Nanny's feelings on the matter, maybe they simply care about women as people much more than the working-class men, and don't feel like going to war against them?
What I found odd about that statistic is I would have assumed only the upper middle classes could cheat with the nanny, given that affording a life in nanny is one of the attributes that defines you as upper middle class.
Well right, but the example (as I understood it) was working-class men telling upper-class women that their husband was going to cheat with the nanny, possibly as a way of degrading the upper-class women for being ugly or career-oriented or not submissive or w/e.
So, working class men projecting their fantasies of what they would do if they were upper class.
There is also the historic example of women working as servants where they became pregnant by the master of the house or one of his sons, and were speedily kicked out. So working class people may have more memory of what the upper classes get away with when it comes to the lower classes.
Being pregnant by or having a child with the master would give the servant soft power to wield in the household. This both dilutes the master's power and creates leverage for instability, especially between the servant and the master's wife. The best remedy is to use hard power to eliminate that threat to stability.
Nothing stops the employer from giving the servant a modest sum to live on or get set up elsewhere, but there's also nothing compelling that and it opens up the possibility that it could be used against them later. The master could always claim the servant was let go because she got knocked up and they don't tolerate that kind of immorality. Who are people going to believe, the master, or the servant who did actually have a child out of wedlock claiming it belongs to the master? Giving her money implies she's actually in the right. If you gave every servant who got pregnant money to live on, servants would go pregnant on purpose.
No, it would not. If you're the master's favourite, it will get you better treatment for a while. But a bastard child is not going to be tolerated. The lady of the house will make sure you disappear. The social mores are that pregnancy outside of wedlock is disgraceful, and even more disgraceful if you run the kind of household where the servants are sluts and slatterns.
Some places and times, if there's enough land to carve off the estate, the master may arrange for the pregnant mistress to be married off so some other man can raise the child in exchange for a farm as dowry. That happened round here in the early decades of the last century.
But pregnant maids who may have been coerced or even raped by the men of the household are not wielding any kind of power. Being dismissed without a reference meant you couldn't get a job elsewhere, and a lot of women fell into poverty and did end up on the streets.
On the other hand, upper middle-class couples are the ones vastly more likely to have nannies, and working class/lower middle-class women are the ones more likely to be nannies/working in some childcare role.
So I think upper class women saying "of course Horace would never with the nanny" may be some wishful thinking, and upper class men saying "of course I would never with the nanny, darling" may be CYA.
Tangential, but class and wealth are not quite isomorphic. The owner of a plumbing company could well be making more money than a consultant at McKinsey, but would probably still be culturally lower class. Movie stars, plausibly also.
(I expect there are also McKinsey consultants who have slept with the nanny, so I'm not engaging with that part of the argument).
Yes, I agree that this is a distinction that matters. Ben Affleck is probably culturally working-class, so it's perhaps less surprising that Ben Affleck slept with the nanny than some guy who went to Harvard on a legacy admission.
On the other hand, Ethan Hawke's father was an actuary and Robin Williams's father was a senior executive in Ford; both apparently enjoyed quite privileged upbringings. It seems reasonable to describe both men as culturally middle-to-upper-class.
> Ben Affleck is probably culturally working-class, so it's perhaps less surprising that Ben Affleck slept with the nanny than some guy who went to Harvard on a legacy admission.
You have a very warped view of classes, culturally or otherwise.
It’s not on the least surprising when Harvard graduates are cads, it’s a solid working class guy like Ben Affleck you’d expect to be decent.
( actually on googling him I see his mother was a Harvard educated teacher, so maybe that explains the blackguardism).
>You have a very warped view of classes, culturally or otherwise.
>It’s not on the least surprising when Harvard graduates are cads, it’s a solid working class guy like Ben Affleck you’d expect to be decent.
My comment was quite clearly made in the context of a discussion about an article in which a woman argued that working-class men think that cheating on your wife with the nanny is inevitable. If my view of classes is "warped", then so are the author of said article.
PEPFAR is great and should continue, but I don't think making claims like this is helpful: "PEPFAR Impact Counter tries to estimate the number of people affected, and says that 13,854 adults and 1,474 infants have already died from this policy."
Unless I'm missing something HIV typically takes years to show any symptoms, much less cause of death.
- "The recovery of the immune system among individuals living with HIV through antiretroviral therapy can take several years. However, this recovery is quickly compromised when the virus begins to replicate in the absence of treatment"
- "Of the infants born with HIV who do not receive treatment, 20% are expected to die within the first three months of life "
But I don't have a good sense of how much these are deaths that have happened, deaths that should be expected in the future based on things that have happened now, vs. the amount of deaths you would get if you suspended PEPFAR by X times the amount of time it's already suspended, divided by X. My guess is some combination of all of these.
Do you actually need to see anybody die in order to produce a reasonable estimate for the number of people who have?
Presumably we know roughly what the mortality rate is in populations supported by PEPFAR, and in populations unsupported by PEPFAR; if the difference in those two mortality rates (multiplied by the population size and by the time since PEPFAR was cancelled) works out at 15000-odd people, that seems like the main thing we need to know?
(In fact, unless I'm missing something too, it seems to me to very much be preferable to work out an estimate in this way rather than waiting for those 15000 people to actually die before starting to try to count them..!)
Yes, I understand that - but if we have very good data about what the mortality rate is going to change *to* (for example because we can see the mortality rates in non-PEPFAR-but-otherwise-demographically-identical populations) then surely we don't need to wait for it to actually change before we produce our estimate for the number of deaths?
(I'm fact, if we're trying to figure out whether this policy change is worth the number of deaths it might cause, I think it should be basically essential to produce the estimate before the mortality rate changes! If we're trying to calculate out whether or not we can afford to help some people it hardly seems useful to only start doing the maths after they're dead!)
Okay, sure, it would technically be more accurate to say "the deaths have been irreversibly set in motion but have not-yet actually occurred". I think "have already died" is a fairly common colloquial way to express that in a morbid-but-pithy sentence, but I can see this being inappropriate in an academic work (or else an indication that maybe the author had the same misunderstanding I had*)
* Your emphasis really helps me there - thanks! - I thought the argument was about how early you can reasonably make the estimate rather than, given the estimate, how early you can say "have died")
It's an awfully specific number as well. Saying "approximately 14,000 adults will die/have died from this policy" doesn't imply as strongly that you know the names and circumstances of each individual who died and can directly confirm that the death was from lack of PEPFAR funds.
I suspect the people who produced that number are very well aware of the difference, and are willing to make it sound more concrete in order to gain that extra leverage.
I don't think they're immoral to do so, but I do think we should discount what they say much more because they are willing to do that.
In some of these cases, I think it’s also likely that a three week gap leads to zero deaths, even if a twelve week gap leads to a large number of deaths.
>Related: this is all fun to think about, but very early polling for the 2028 Democratic primary suggests that by far the #1 candidate is . . . Kamala Harris at 37%, beating Mayor Pete, Gavin, and AOC with 11%, 9%, and 7% respectively. I know you’re not supposed to take early polls like this seriously in terms of who will actually win, but can you take them seriously as a guide to whether people have learned any lessons / no longer love losing?
Take it as a lesson on why first past the post voting is absurd, and should be abolished!
Harris is winning that poll is indeed on name recognition alone, which is how a lot of people win first past the post elections with lots of candidates (plausibly how Trump gained momentum in his fist primary, for example).
Try again using Approval voting, you may get very different answers.
Doesn't high name recognition also lend an advantage in approval voting? While some people do just vote at random, I think in general a voter would be less likely to approve a candidate they don't know well.
First of all because if you approve one person because you know their name and 5 other people because they seem alright after you listened to them talk for 5 ins (o w/e), they all get equal support form that. Vs voting for the one person with the big name and everyone else gets zero support.
But, second and perhaps more importantly, because it overcomes the issue of strategic voting through Schelling points. A big reason that many people support high name-recognition candidates is because they expect less engaged voters to know who that person is and rally behind them, so it feels safer to vote for the well-known person to avoid intra-party infighting and lost momentum. With Approval you can still approve that person in case that is what ends up happening, but also approve everyone else you like and give them a much better chance of competing.
I am not that immersed in American politics so I'm actually curious why they don't just fucking run Hillary again. Harris now has the exact same record as her against trump!
“They” don’t run anyone. The person themself has to enter the race. Everyone who says “why don’t they just…” doesn’t understand how American political parties work.
FPTP has the advantage that it is simple to administer and simple to understand. We already saw how much of a shit show things can be with people not understanding how elections operate. It would be even worse if you use some fancy voting system that doesn't translate to results in an obvious way.
Approval is about as simple and easy to administer.
'Vote for everyone you like, most votes wins' is actually fewer words than 'vote for the person you like most, most votes wins. And you can use existing ballot and voting machines with no overhaul, just let people fill in multiple bubbles.
Re: 22, I don't share your views on DEI but I still struggle to steelman them into a view of their impact on federal workers that produces anything close to the harm of the doge cuts. Just look at the Ted Cruz survey you posted about -- some studies you don't like but they were a fraction of a fraction.
It seems to mee that the FAA example is at an extreme and was recognized and addressed by the system eventually. Also there is lots of evidence of systemic discrimination against non-white applicants in many fields (not specifically the government as far as I know) so DEI policies are doing some good even if in some cases they are also causing harm. Even if 25 percent of hires are "impacted" by DEI considerations, Musk is talking about straight up firing far more than 25 percent of federal workers, though he hasn't quite gotten there yet.
I think it's implausible that removing discrimination helps more than affirmative action hurts. The most easily measurable thing here is college admissions, and we know that admitted minorities got much worse SAT scores and other measurables. This is also my impression of the Get More Women In Tech movement. It's possible government is an exception, but I haven't heard any good arguments why that would be.
I think lowering the quality of workers probably hurts more than lowering the quantity - if one general makes bad decisions, that's plausibly worse than having 25% fewer soldiers. But I agree that if Musk fires "far" more than 25% of federal workers then it will become implausible that DEI had a larger effect.
I think both Musk and DEI will have their largest effect in very smart people with other options realizing that it would be masochistic to go into public service, and sticking with Wall Street or something instead. I think government needs some of those people to function well and could barely do a better job kicking them out if they tried.
I guess I have a hard time seeing how you get from *any* of these things to the chilling impact of putting forward indiscriminate cuts as a possibility. Even if they stop tomorrow and don't fire anyone else, they've already decimated the federal hiring pool for many years, until some Constitutional change can credibly promise that this sort of thing won't strike again.
I don't follow. The boss of Google can lay off any Google employee at any time, but Google can still attract talent. My impression is that talented people are willing to work at places where they might be laid off as long as the chance is low. I've heard claims that DOGE wants to fire ~10%. Suppose that every Republican president does something like this and Republicans win half the time - a 10% chance of getting laid off per eight years doesn't seem that different from private industry.
There is the obvious difference that Google pays a lot more (in dollars) than the government does. In the past the government was arguably paying partly in security, but the government total compensation package has just received a drastic cut. You would expect that to hurt recruitment, unless they up their pay in actual dollars to compensate.
Along with the point about government jobs compensation package including less money but more security, the general assumption at private companies is that during layoffs management attempts to dismiss only their poorly-performing employees. So you can avoid being laid off by working hard and being good at your job. Obviously in the real world no one has solved the problem of accurately measuring every employee's value/quality, but at least they try.
The DOGE cuts have conspicuously not even attempted to fire bad employees and retain good ones. Federal government departments do have HR and managers and annual reviews and such just like private companies, and one would think a bunch of programmers whose skills are sufficient to fiddle with the US Treasury's money disbursing database could also analyze the HR databases, but here we are...
Google has in fact become, from my social circle's anecdata, a significantly less desirable place to work since they started doing mass layoffs in 2023 for the first time. People were particularly mad about the lack of transparency around the layoff decision making process and the lack of obvious correlation with performance. It wasn't as bad as DOGE in those respects, but it reflected an arrogance towards the rank and file that seems unfortunately to have gotten more common even among non-insane tech executives over the last 5-7 years. The rank and file are not blameless in this-- the cartoonishly dumb woke techie activism of the late 2010s rivaled that in universities-- but the phenomenon is nonetheless very bad and likely has eroded morale and productivity in much of the industry.
This misunderstands the (to date) compensation structure of Federal employees, including some of the uniformed service members effected by DOGE decisions and DEI purges.
Generally, Federal compensation is heavy on non-salary compensation. Historically, this has included defined benefits pensions which vest after long periods of service, but that's mostly been phased out these days in favor of other things. Recently, this has also included flexible telework options for most employees. You also get a measure of job security not enjoyed by most private sector folks and often had opportunities to directly effect unique missions that weren't accessible in the private sector. In exchange, you get paid less than you would in the private sector. In some areas, this difference in pay is so large that the Federal government has struggled mightily to recruit talented individuals.
There was an understanding (varying by degree of how politically-facing your office was) that you would be insulated from the day-to-day shifting of the political winds. For example, the Commissioner of BLS might have different priorities depending on who appointed them, but those changes were mostly around the edges, and the core mission of collecting and publishing credible data on the market wouldn't change.
And for the most capable employees, a crucial part of the compensation package was the mission. Private companies can also have appealing missions, but there's a lot of pro-social work that can only be done at the Federal level, and for many Federal employees (especially the best, who could easily find other, more remunerative jobs in the private sector), that was the primary appeal. Or at least a key component that compensated for some of the other, less enticing aspects of working for the Federal Government.
These past two months have struck directly at those non-pecuniary aspects of the compensation package.. Apparently, you can be fired capriciously by some rando Presidential Advisor. And you'd better get damned political, because the current party thinks every Federal job is a political appointment, and damn your mission.
The approach DOGE is taking is to eliminate the professional bureaucracy that served the missions of the agencies they were a part of. It takes as a starting point that bureaucrats should serve the President first and their agency (and their agency's missions) second. This is part and parcel with the a "unitary executive" legal theory the current administration is aggressively pushing. I have no opinion on its merit w.r.t. the Constitution, but its effect on the professional bureaucracy and the independent agencies they support will be devastating.
Surprised by this take. I'd much rather have a fully staffed IRS than a 80% staffed IRS with slightly more competent people on average.
Also, talented people who choose to take government jobs instead of industry when both are available do so because of stability. These people are very risk averse and defaulting on them once (or just laying off anyone) is really bad for attracting risk-averse and talented people in the future.
I wonder if that's why government norms seem to be way too risk-averse to be compatible with sanity. (Classic examples: FDA approval, Covid challenge trials)
These are decent examples, but for a lot of function of government, risk-aversion is a good thing. The worst outcome of taking risks as a private company is (usually) that the company fails. The worst outcomes of taking risks as a government responsible for hundreds of millions of people make entire chapters in history books.
I’m having a bit of trouble following the details, but it looks like it was started in 2014 and stopped in 2016, though there was some long-lasting damage.
He doesn't seem to think that the law fixed it, since he goes on to say "The Trump administration missed it, too, for a term, and it’s likely most officials simply didn’t hear about it through the first few years of the Biden administration." and "I badly wanted the Biden admin to rise to the occasion, speak plainly and frankly in response to the scandal, and defuse this time bomb." and "the Biden admin declined to fix it" and "Democrats had a chance to clean it up, and they didn’t", all of which are about events after the law was passed. So if Trace is your source, he's also the one saying that the 2016 law did not fix things.
The 2016 law did not provide recompense to the affected employees, but it did eliminate the hiring practice going forward. So it all depends on what we mean by "addressed by the system." There's an ongoing class action lawsuit seeking that recompense (and we only know the incredible details here because they pulled those details out during discoveries over the last decade.)
No, he's arguing that DEI is good because it removes discrimination, I'm arguing that the effect of discrimination on productivity in negligible because the majority of competent applicants are whites, due to demographics and... other reasons. It's basically the same argument Scott is making above.
There is also lots of evidence of systemic discrimination against men, favouring women, and yet we were constantly told the opposite, and institutions of power pervasively engaged in overt discrimination (like women only fellowships) in addition to the covert one. Somehow gender discrimination gets less attention than racial one, but there's a solid case for men to be angry.
Taking that number at face value, is 25% of federal hirings and promotions over a four-year period really all that big? If (at a wild guess) the average person who works for the federal government at all in their life puts in 20 years in total, that's only ~5% of hirings that would be affected (I'm less sure how to guestimate promotions).
Meanwhile "impacted by DEI concerns" seems like quite a low bar. How much impact are we talking about here? Intuitively, it seems like this would be the biggest issue in cases where the qualified applicant pool is very small (so the value-over-replacement of the best candidate is likely to be high), which is *not* my impression of what most important federal jobs are like.
DEI doesn’t seem to be particularly harmful to private industry. McKinsey has put out four reports showing that companies with more diversity on their boards and in top management perform better. The fact that they got similar results in different years, and while more than doubling their sample size (so they weren’t just repeatedly looking at the same companies) suggests that this is not just a statistical fluke.
I found the following article that pushes back against this:
The article correctly points out that correlation does prove causation. It also points to academic research which shows that “the relationship between board gender diversity and company performance is either non-exist (effectively zero) or very weakly positive.” It cites a finding of a minimal (though statistically significant) correlation between gender diversity in top management and corporate performance of 0.03.
Perhaps DEI program harm corporations in way that these studies can’t pick up on because the studies only look at diversity in top level management and on boards of directors, ignoring lower level employees. It is also possible that a DEI program that is harmless in a corporate setting will have a different effect in a government setting. Still, I’d need a bit more than anecdotal evidence to conclude that DIE programs were doing major harm to government efficiency.
Ah yes, McKinsey, known for their dedication to unbiased knowledge-seeking and absolutely not for shitting out 100 page slide decks supporting literally any position a client is willing to pay for.
This paper, contra McKinsey, doesn’t find a statistically significant benefit to diversity in top management. I’m inclined to trust it more than McKinsey.
The question I was considering was not whether DEI helps financial performance, but whether it hurts it. This paper doesn’t find a statistically significant effect, which leaves open the possibility of a negative effect that is too small to detect without a larger sample size. However, the Wharton piece references a metastudy that did find a very small but statistically significant positive correlation. So I’d say the evidence still points towards no negative effect.
Thank you: I'm not a *fan* of a lot of the DEI stuff, except to the extent it accelerated the trend of people not being openly racist and sexist on the workplace (still a thing ten years ago), but the harms seem wildly overstated outside of academia.
... and people don't see how much DEI in academia is by gender and how little is by race, because most people's experience of academia (if they have any) is centered at the undergraduate level, and thus gives much more weight to undergraduate admissions than to anything else.
In a series of very influential studies, McKinsey (2015; 2018; 2020; 2023) reports finding statistically significant positive relations between the industry-adjusted earnings before interest and taxes margins of global McKinsey-chosen sets of large public firms and the racial/ethnic diversity of their executives. However, when we revisit McKinsey’s tests using data for firms in the publicly observable S&P 500® as of 12/31/2019, we do not find statistically significant relations between McKinsey’s inverse normalized Herfindahl-Hirschman measures of executive racial/ethnic diversity at mid-2020 and either industry-adjusted earnings before interest and taxes margin or industry-adjusted sales growth, gross margin, return on assets, return on equity, and total shareholder return over the prior five years 2015–2019. Combined with the erroneous reverse-causality nature of McKinsey’s tests, our inability to quasi-replicate their results suggests that despite the imprimatur given to McKinsey’s studies, they should not be relied on to support the view that US publicly traded firms can expect to deliver improved financial performance if they increase the racial/ethnic diversity of their executives.
You have no evidence for that 25% number and even if it was true, do you have evidence that the people hired were of less quality? If you have two candidates of equal quality, and you hire the black one because of DEI, thats a "DEI hire" but it doesn't impact the quality of work.
I suggest reading the TOOW review. "Only when two candidates are exactly equal" was always a fig leaf - the law demanded that companies massively change the ethnic composition of their workforce in a way that they couldn't do while only hiring exactly equal candidates unless you grant implausible assumptions about how often there were two exactly equal candidates of different races applying for the same job. This is also not the experience of hiring managers I've read commentary from.
I read the review when it first came out and skimmed it again just now to refresh my memory. Nothing in the review provides any actual evidence on the impact of supposed DEI hiring on the effectiveness of the federal workforce.
Data that I would like to see:
How many people got jobs because of DEI policies they wouldn't have gotten without DEI policies.
The people hired through DEI policies are less effective than those not hired by DEI policies
Heck, I would take any objective data on the federal workforce being less effective now than in the past!
"law demanded that companies massively change the ethnic composition of their workforce"
I thought we were talking about the federal government here?
Anyway, which law specifically? "read the TOOW review isn't a good answer, there are lots of civil rights and EEOC laws. TOOW presents a long chain of inferences and assumptions about what may have happened. But where is the evidence that it actually has?
If DEI has been such a horrible policy, then it should be trivial to present data about it. Yet I haven't seen any! And I ask for it in these threads all the time!
"(~25%) of federal hirings and promotions were impacted by DEI concerns" can only be justified by an absurdly weak interpretation of "impacted" or an absurdly broad interpretation of "DEI concerns".
The FAA scandal is an excellent example because it is simultaneously (1) egregious, (2) exceptional, and (3) per TW's own assessment, had minimal overall impact on the quality of ATCs who made it out of the pipeline. Revelations like that should cause you to believe that DEI has relatively small effects on hiring and promotion decisions, not large ones.
I had the same reaction to this bit, while generally agreeing with Scott's pessimism re: the admin's cost-cutting approach. Specifically:
"I think the main effect will be saving ~1% of the budget at the cost of causing so much chaos and misery for government employees that everybody who can get a job in the private sector leaves and we’re left with an extremely low-quality government workforce. I freely admit that DEI also did this, I just think that two rounds of decimating state capacity and purging high-IQ civil servants is worse than one round."
This assumes that DEI practices left us with "an extremely low-quality government workforce," which seems to me like an empirical claim not really supported in Scott's responses here besides anecdotes on hiring from the comments on his Hanania book review and an analogy to college admissions and SATs. Is there output-based evidence suggesting this is true? May be I'm unduly assuming government work is an exception of some sort.
FWIW, from what I've seen and heard (D.C. lawyer), in the legal field you're already choosing from a slimmer talent pool with a GOP admin because lawyers trend liberal and many don't want to work for it (this may be a Trump era development, not sure). I suspect the Trump 2.0 politicization of the administrative state likely will exacerbate this problem. Plus we've already seen them drive away even very conservative senior staff with tomfoolery (see the Eric Adams deal and subsequent resignations). So firing experienced staff otherwise willing to stick around may be especially harmful for competence in legal positions.
The SC rules that only Trump personally is above the law (in the course of performing his official duties). This doesn't apply to subordinates and you can only pardon so many people before the Presidential pardon will be taken away by Congress. (I wouldn't be surprised if that happens in the next few years, in any case.) It's also a problem that the pardon only plus to violations of Federal law.
Biden could have become the ultimate based president if he had pulled out a gun and shot Trump during their post-election meeting. According to the insane logic of the Supreme court, it would have been legal or at least in a mere legal Gray area. Maybe Trump can use that one weird trick on the president elect if the Dems win in 2028.
> This doesn't apply to subordinates and you can only pardon so many people before the Presidential pardon will be taken away by Congress.
...The congress that's controlled by Republicans that stood behind Trump after Jan 6 instead of throwing him under the bus? That continues to stand behind Trump as he commits an ideological purge of federal agencies and destroys old alliances to ally with a dictatorship? These people are absolutely in it for the long-haul. As for the mid-terms... there are an endless amount of ways to deal with that when the law no longer matters.
You're fantasizing about some kind of revolution. But Trump isn't a revolutionary figure and the Republican party is not a vehicle for revolution. The Republican party is fundamentally conservative. It's not going to allow Trump to just sweep away all laws in order to do a revolutionary overhaul of the country. That's just a fantasy. Even Trump supporters have their limits. Notice how they boo Trump when he talks about the beautiful vaccine that he created. Americans believe in the law, and Republican voters care about the rule of law as much as anyone.
Hell, disorder and lawlessness is a big part of the reason why they have rejected the democrats. A majority think that January 6th was a false flag executed by antifa members. You've been reading a lot of liberals complain about how Trump is a revolutionary Maoist who is going to totally upend everything, and I'm afraid you've gotten your hopes up. Are you even American? It really sounds like you interface with this country indirectly, not directly.
It's easy to stay detached when you have nothing to gain and nothing to lose. Regardless, I doubt I'll be able to convince you, and there isn't much point in doing so in the first place when there's absolutely nothing you can do to stop this. So please... just enjoy the show.
I have seen no evidence that "DEI hiring" has caused any impact on federal workforce effectiveness. Its astounding to me that Scott and so many "rationalists" just take these claims as true without any evidence of it.
Is AIs' refusal to talk about topics like race/IQ not just the same kind of "woke" opinion that's explained by the liberal/conservative divide in training data?
It's hard to get refusal to talk out of training data, because AIs are specifically trained to want to answer all questions, so you have to specifically train them out of it. If they'd just given a typical liberal answer on race/IQ, I would believe it was just training data. I think AI companies chose not to go that route because you could always prompt them differently to give a non-liberal answer, unless they were trained otherwise.
Whatever podcast is so bleak. Just this dude trying to catch out girls who aren't even old enough to drink in a logical inconsistency so he can feel smug about them because he knows he'll never get to fuck them. Although admittedly the Gorlock memes were pretty funny.
I've never heard of this podcast before and can scarcely imagine anything I might possibly care to watch less (also: since when did podcasts become something one *watches*?) - but if 'catching beautiful women in logical inconsistencies during debates' is this geezer's kink, who are we to judge?
Andrew Wilson is...certainly something. My cursory understanding is that he also argues against democracy with randoms brought onto these podcasts. It's the most annoying form of engagement, but it gets clicks, so Whatever, I guess.
I'm not sure that Private Eye does have that great a record. They throw a lot of mud on the wall, and some of it sticks, but plenty of it is quietly forgotten.
They went all in on supporting Andrew Wakefield, the doctor who linked vaccines to autism.
They do some excellent journalism and book reviews. But their editor, Ian Hislop, is a fanatical Europhiliac, who never misses an opportunity to knock Brexit! That spoils the magazine for me.
The biggest piece of evidence is the fact that she was found guilty in court.
All of the big expert claims and long debunkings are classic "man of one study", although it might be better termed man of one claim.
They go through the available public evidence, find a specific medical or other technical claim, then go expert shopping to find a group of experts who disagree (which as you no doubt know for medical claims, is never very difficult). Then, having found enough doubt to cast on this claim, hey presto, Letby must be innocent.
It might well be that the pro-Letby experts are entirely correct in some cases. But, no reference is ever made to all the other details of the case which were used to convict Letby. There was a lot of traditional police work that went into her conviction that is ignored
This was an expert they got well after the initial case, when they presented other experts, and after other "expert panels" ruling on other minutiae had already been presented and failed to make the desired impact.
It's hardly minutiae when the expert wrote the paper the prosecution used to propose a murder method. And when that same expert says they misused his paper and the baby could not have died in the way the prosecution said.
I personally believe that she is guilty but retain 5% probability that the evidence is faulty. My suggestion would be instead of guilty or not introduce another category – probably guilty – which is not enough to jail someone but enough to prevent to working with children.
I have to ask then why was Letby charged. Was she made a scapegoat for the failings of the hospital care? Was it just a coincidence that she started working on wards where the mortality rate went up? Bad luck? Why her and not someone else?
There have been enough killer nurses that I suppose there's a certain amount of hysteria at work. And Letby was just in the wrong place at the wrong time. Assuming she's innocent, which I believe, although I'm no expert on the case.
If Letby is exonerated on a "reasonable doubt" that doesn't prevent the NHS refusing to employ her, her professional body refusing to let her practice or Child Protection measures being taken against her.
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
The case might be unstable in a few of the medical areas, but it seems unlikely that this is all just an unfortunate coincidence for Ms. Letby.
Re: gondolas. I spent a season as a ski bum pre-career, with the interesting wrinkle that I specifically worked on what was at the time billed as the only gondola-based municipal mass transit in the US. AMA I guess, and some thoughts:
- Gondolas obviously have a huge advantage in mountainous terrain where other mass transit is a non-starter. I'm not sure the economics will pencil out otherwise given the fairly limited throughput, though some designs can get to some fairly impressive capacities.
- The individualization of compartments is both a blessing and a curse, overall I like the trade. The consistency of service and increased privacy is quite nice off of peak hours. People can and do make messes in the cabins, with vomit being by far the most common (same issue as Uber/Lyft, for people coming back from the bar). Hard drug use was nonexistent, hotboxing was common - though that might have been an artifact of this being Colorado in 2014. It's trivial to take an individual cabin out of service for cleaning with minimal disruption.
- This is a high-contact service position; operators *have* to know the local PD has their back. Standard practice for egregiously disruptive members of the public was to let them on the next cabin, then punch in a stop once they're in the air until PD can meet them at the next station. Striking an operator is like 3 different felonies, DON'T do it.
- You will need to pay operators above local prevailing wage if you want folks to pass a drug test and get that sweet Federal/State funding. They *will* fail if you retest often though, see hotboxing note above.
- All the various issues of "how do we accommodate people with XYZ disability" have been solved, but in most cases it results in a 30-60 second delay for everyone on the line.
Re #6 l, the main problem is how little signal you could ever extract from a resume. I/O psych research generally favors intelligence tests, structured interviews (although some researchers still apparently think unstructured interviews can be solid), realistic tests of job skills, a small number of personality dimensions, and experience (but only in the 0-5 year range), as good predictors. Educational history is weak, references are weak, experience over 5 years is weak. The only clear signals you can pull from a resume are "does this person work in this field?" and "have they for 5 years?"
Beyond that I'd be really suspicious of any AI system that supposedly does a good job; it's probably just figuring out how to reflect the biases of the people using it.
Re 3 - your take is mostly true. My career background is in engineering at large ISPs. so I have significant understanding here, although there are people in the industry that have better grasp on the regulatory side. A caveat is that I've never worked at a residential provider, but I am an industry expert on backbone topology.
Without getting too deep in the weeds:
1. At the time of the second big push for Net Neutrality (roughly 2014-2017), most ISPs backbones were on speeds and/or protocols that could be considered two grades behind where things are currently and had been relatively static on this front since roughly 2006. The 2014-2017 period coincided with when adoption/utilization of streaming services, which is unequivocally the biggest driver of Internet bandwidth, really started to uptick.
2. The content providers, even those that were early adopters of CDN approaches, realized around this time that they couldn't continue to grow if user experience over the insufficient average home Internet connection of the time continued to be poor. In addition to pressuring the major providers to invest in their networks (often by spending money directly with them for this purpose), the largest content providers also radically rearchitected their own environments in order to reduce the impact that a user's ISP architecture would have on end user performance.
3. Continued investment of federal funds into broadband improvement - while there is plenty to be aggrieved about with the implementation of nearly all of these programs - has sustained continued improvement in the average home Internet user's experience through all subsequent Presidential and, therefore, FCC administrations.
Notice I haven't talked about net neutrality or the regulatory environments at all.
While the previous statements are facts, the following statements are (I believe to be highly informed) opinion:
1. Some of the things Net Neutrality proponents warned against are not viable, because (most) ISPs are capital-constrained and extremely slow to innovate, even (often especially) when the larger parent companies are massive money-printing machines, and would require them to be nimble and agile.
2. Some of the things Net Neutrality proponents warned against are viable and definitely deployed by many ISPs but have little to no adverse impact to end users. Zero rating certain traffic is a good example of this.
3. Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
4. Net Neutrality is actually a red herring - on the Internet, treating traffic equally and treating traffic neutrally are different and both principles violate reasonable traffic management practices. The main reason for this is that Internet backbone routers are essentially armed to make their own decisions, and while the parameters to make those decisions can be tightly controlled, the end result of those changes can be variable, even when done by seasoned veterans in well-controlled circumstances (and often neither of those things are the case). Lack of customer choice due to local monopolies or duopolies is a real problem (which emergent providers and technologies are slowly addressing) but Net Neutrality is a theoretical argument, and customers care about user experience not philosophy.
5. This is slightly less informed opinion, as I have never worked for a residential ISP, but I don't believe that (most) ISPs are agile enough to change their traffic management practices based on regulatory environments that can potentially change every four years federally (let alone state-by-state environments) and therefore not responsive to it at all.
The zero rating thing is the reason why in my own (admittedly much less technically informed opinion), that I said I thought lack of net neutrality is probably actually having small, negative impacts.
Zero rating specifically benefits large incumbents who can make those kinds of deals. It therefore dis-incentivizes new entrants.
Theoretically, we should expect a reduction in competition because the existence of zero rating presents a new barrier to entry/competitive friction.
As I also mentioned, I expect this effect to be relatively small and hard to see, definitely much less severe than what the doomers were saying at the time. "
I actually agree with you about zero rating in theory. In practice, however, I believe it to be a non-issue. My guiding principle is always "is the end user experience adversely impacted?" and as zero rating is a commercial practice and not a technical one, it requires end users to make different utilization decisions specifically on account of some content being zero rated in order for the answer to that question to be yes. Industry data that I've seen consistently shows that this isn't happening. That said I firmly believe that the increase in utilization caps for home Internet and mobile plans, along with topology changes on the content side, have made zero rating more of a non-issue for end users than anything else - in other words, traffic not counting for customer utilization doesn't really matter in that environment. I'd like to see some specific data for users that are on more tightly constrained plans but it's such a small portion of the overall user base at this point that I'm unclear if the data would be meaningful.
The other issue you point out is creation of a barrier of entry on the content side. Again, on paper, I agree, but in practice there are simply much bigger barriers of entry. On paper "start up creates competitive search engine to Google and gets blocked out because end users can use Google for 'free' but not the start up because they can't afford to pay off Verizon" sounds bad, but the billions of dollars the start up would need to create infrastructure that can support creating a product competitive with Google's is orders of magnitude larger than whatever deal would be struck with a carrier. Moreover, so much of this type of new entrant lives in hyperscale cloud infrastructure that this isn't really where zero rating comes into play anyway (in part because the level of granularity required for a carrier to differentiate between traffic sourced from, for example, within the same AWS-owned IP block, lacks a cost benefit), it's for the "competitor to Disney+" or "competitor to Spotify", services that source much more data in aggregate and in wholly different ways than a more static website, and there a whole boatload of bigger reasons why competing with those services as a start up are extremely challenging, primarily access to the source content.
> 3. Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
1. Providers that deliver private traffic (which to be clear is a completely separate service from the public Internet and therefore outside the scope of any Net Neutrality type of regulations) on the same backbone trunks as public traffic *always* prioritize the private traffic, by definition and necessity. This is completely legitimate but many vocal NN proponents have conflated this with the end users of private services and carriers illegitimately working together when in fact this is a *completely separate service*.
2. At the same time, the Internet is a best effort service, which is true by design and definition. Part of what this means is, by definition, prioritization of Internet traffic (which can occur under an umbrella of methods typically referred to as Class of Service or CoS) generally doesn't work. The full version answer here is much more nuanced, but it doesn't work primarily because carriers don't respect prioritization markings delivered from other carriers (and therefore their directly/indirectly connected end users) unless they pay each other for it, which does not happen for public Internet traffic. I fully realize that this somewhat recursive and potentially tenuous, but at the same the entire Internet effectively works because of the same game theory, so isn't particularly worrisome.
Now, what providers *can* do, and *did* do, and which was a proximal cause of the second wave of Net Neutrality conversations in the 2008-10 era, was completely deprioritize traffic coming from other providers in which there was some sort of business dispute. Google "Cogent Hurricane Electric cake" if you're so inclined. This is dissimilar in cause from traffic prioritization but is similar in effect, at least for the directly/indirectly attached customers of those providers. This sort of stuff still happens from time to time behind the scenes (and not to pick on Cogent, I have some friends there, but 98% of the time it involves Cogent) but it has essentially no impact on end user experience now. Why? Because the big content companies, who were really the root cause of these types of disputes most of the time (i.e. Hurricane Electric was getting paid by Google as their provider, Cogent wasn't, HE was dumping a bunch of traffic to Cogent's network that flowed to end users that Cogent claimed they couldn't afford to upgrade their network to handle because Google wasn't paying them, etc.) completely changed the way they delivered content. The specific details aren't important, but the ubiquity of CDNs, local caching (in the content sense). edge compute and similar technologies, have made these types of challenges immaterial. It's probably also worth pointing out that the effect of this blunt-force deprioritization lacks granularity and it is a very different technical problem, but really the content providers have made it such that the end user impact in these situations is negligible to non-existent.
> Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
Can you explain why this is something we shouldn't worry about? If there are technical limitations I'd love to understand.
22: Maybe in some places DEI had the same effect as DOGE...but I think you are wildly overstating it. In my experience, DEI consisted of sitting through one EEO training and one sexual harrassment training per year. DOGE is...significantly more damaging than that. Do I think 'send me an email with five things you did' is fascism? No, but I think pretending 'send me an email with five things you did, tomorrow by midnight, or you're fired,' by someone with...what governmental position exactly? Is in any way compliant with civil service protections, proper division of power, or general morality (or that it's something anyone's going to do in five minutes--your job's on the line, but it'll only take five minutes, is not an internally consistent position for anyone who has ever met a person) is bullshit.
24: This is sort of correct, but mostly irrelevant/bad. NEPA didn't go away, it's a statute, everyone still has to comply, all they've done is remove the CEQ regulations that provide guidance on how to do so. Agencies have their own regulations...which are often dependent on CEQ regulations. The agency regulations are what contain the categorical exclusions that allow you not to have to do an environmental assessment for...everything. It's not at all clear at this point whether those regulations are still valid and the desire to mess with them betrays a failure to understand that that will open them up to litigation (they're generally too old to be sued over on their own, rather than as applied to individual actions).
More broadly, the NEPA regulations had some good and some bad (hey, can we adopt other agencies CATEX's? Probably not anymore, since that was in CEQ's regulations! Great, thanks). Removing them removes basically no requirements and I think some of the climate change stuff just puts agencies on a collision course with courts which are going to continue to maintain that the requirement to evaluate environmental effects obviously includes climate change.
Subverting procedures and the rule of law and setting up a system where only the personal opinion of the Glorious Leader matters are textbook fascist moves.
By that definition both Hitler and Stalin were fascists, which is interesting since they considered each other mortal ideological foes and killed tens of millions of each others’ citizens to prove the point.
I think the word you are looking for is “autocratic”.
I recently came across a funny/true/interesting statment about semantic drift:
At first it's a mistake, then it's not!
I think we are pretty clearly somewhere on that curve for "fascist". The way it is currently generally used in the culture absolutely would have been wrong 50+ years ago, and probably also would have been wrong 10-20 years ago. Is it _still_ wrong or has the definition changed? When will we know?
“Fascist” is the insult the left applies to people doing authoritarian things they don’t like. “Marxist/socialist” is the insult the right applies to people doing authoritarian things they don’t like. Both sides call people doing authoritarian things they *do* like “strong leadership”.
Generally I’m against obfuscating historically meaningful terminology for the sake of punching up current day political shit-flinging.
I am not sure it was ever that useful, this is an Orwell article from 1944 ans it seems mainly a left wing insult back then.
"It will be seen that, as used, the word ‘Fascism’ is almost entirely meaningless. In conversation, of course, it is used even more wildly than in print. I have heard it applied to farmers, shopkeepers, Social Credit, corporal punishment, fox-hunting, bull-fighting, the 1922 Committee, the 1941 Committee, Kipling, Gandhi, Chiang Kai-Shek, homosexuality, Priestley's broadcasts, Youth Hostels, astrology, women, dogs and I do not know what else.
Yet underneath all this mess there does lie a kind of buried meaning. To begin with, it is clear that there are very great differences, some of them easy to point out and not easy to explain away, between the régimes called Fascist and those called democratic. Secondly, if ‘Fascist’ means ‘in sympathy with Hitler’, some of the accusations I have listed above are obviously very much more justified than others. Thirdly, even the people who recklessly fling the word ‘Fascist’ in every direction attach at any rate an emotional significance to it. By ‘Fascism’ they mean, roughly speaking, something cruel, unscrupulous, arrogant, obscurantist, anti-liberal and anti-working-class. Except for the relatively small number of Fascist sympathizers, almost any English person would accept ‘bully’ as a synonym for ‘Fascist’. That is about as near to a definition as this much-abused word has come.
But Fascism is also a political and economic system. Why, then, cannot we have a clear and generally accepted definition of it? Alas! we shall not get one — not yet, anyway. To say why would take too long, but basically it is because it is impossible to define Fascism satisfactorily without making admissions which neither the Fascists themselves, nor the Conservatives, nor Socialists of any colour, are willing to make. All one can do for the moment is to use the word with a certain amount of circumspection and not, as is usually done, degrade it to the level of a swearword."
No, because I don’t know enough about it to even begin to compare them. It was just the idea that if two people share the same political philosophy they can’t get into a fight with each other about something that I was questioning.
Fair, but I was not saying “fascists could never fight each other”, I was saying specifically that Hitler and Stalin agreed that they were leading two fundamentally incompatible ideological systems that were destined to duke it out for world supremacy. And that incompatibility was deeper than just “we’re both nationalists and only one nation can win”.
The Soviet Union had some common ideological point with fascism, yes, including the obsession with a both pathetically weak and yet deceptively strong and omnipresent enemy, but *not* the glorification of violence for its own sake and of men as aggressive warriors.
A definition of fascism that includes the system that both fascists and nonfascists of the time period agreed was the greatest ideological opponent of fascism is clearly missing something.
Horseshoe theory has a long history. Stalin's version of the Soviet Union moved a long way towards the prototypical fascist regimes from orthodox Marxism in style and practices. Primarily the hollowing-out of internationalism to replace it with barely disguised nationalism and the outward-facing militarism.
I think *everybody* moved toward nationalism and militarism during WWII, for not particularly surprising reasons. The idea that nations should *always* operate that way was a key point of Fascism.
Now, Stalin plus nationalism/militarism looks a bit more like just straight fascism than FDR or Churchill making the same moves, but again the key connection is autocracy and totalitarianism, not fascism per se.
Fascism is when you fire government workers. The more government workers you fire, the more fascist something is.
The Nazis were famous for being against big government. Everyone knows the first thing they did in office was slash budgets and personnel. All of the infamous stormtroopers were private sector employees and unpaid volunteers.
"The first mass privatization of state property occurred in Nazi Germany between 1933 and 1937: "It is a fact that the government of the National Socialist Party sold off public ownership in several state-owned firms in the middle of the 1930s. The firms belonged to a wide range of sectors: steel, mining, banking, local public utilities, shipyard, ship-lines, railways, etc. In addition to this, delivery of some public services produced by public administrations prior to the 1930s, especially social services and services related to work, was transferred to the private sector, mainly to several organizations within the Nazi Party."[14]"
But also, this is just an incredibly, incredibly weak argument in context. The thing that was being described as "fascist" was not "mass indiscriminate firing of government workers." It was "choosing which government workers to keep based solely on the whim of a strong, central leader." That's a rather different sort of animal.
The Nazis were pro privatization, yes. Is that the main quality that defined their government, that people actually remember and use when comparing them to other governments? Including other governments that are pro privatization?
>But also, this is just an incredibly, incredibly weak argument in context. The thing that was being described as "fascist" was not "mass indiscriminate firing of government workers." It was "choosing which government workers to keep based solely on the whim of a strong, central leader."
It is a loyalty test. In order to result in any sort of fascist government that will do the bad things that people whine about, it would still need to be hiring a bunch more loyal employees than will exist after the mass firings.
You don't necessarily need to fire and replace everyone, just install loyalists in places where they can control key functions and can punish people who step out of line. Set up your own power structure rather than trying to replace everyone in the existing structure.
On a related note, Elon wants to have a "DOGE Team Lead" assigned to every government department with authority over all hiring and firing for that department.
Federal employees are used to having a lot of leeway to obstruct any presidential agenda that doesn't fit with what the agency was already doing. The feds were getting in the way constantly during Trump's first term. You would need to punish most of them - so it's easier to just do mass firings.
"Is that the main quality that defined their government..."
Not sure why you're asking me questions about the shortcomings of your own argument. If you suggest that some feature precludes a government being similar to the Nazis, when the Nazi regime in fact had that feature the question of "how central was that feature to Nazism" is one that YOU brought into play. If it turns out that it's irrelevant, then your original argument was irrelevant. Can't have it both ways.
"It is a loyalty test. In order to result in any sort of fascist government that will do the bad things that people whine about, it would still need to be hiring a bunch more loyal employees than will exist after the mass firings."
So you agree that loyalty-based mass firing *would* be a pretty necessary first step in installing an effective fascist government, then? Because based on that, it sounds like the complaints and concerns are actually quite reasonable (unless one somehow assumes that hiring people is something Trump's administration is literally incapable of).
"Not sure why you're asking me questions about the shortcomings of your own argument. If you suggest that some feature precludes a government being similar to the Nazis, when the Nazi regime in fact had that feature the question of "how central was that feature to Nazism" is one that YOU brought into play."
The Nazis privatized, but they didn't do mass firings to clear out the security state. They were known for massacres. See the difference?
"So you agree that loyalty-based mass firing *would* be a pretty necessary first step in installing an effective fascist government, then? Because based on that, it sounds like the complaints and concerns are actually quite reasonable (unless one somehow assumes that hiring people is something Trump's administration is literally incapable of)."
It should be telling that the uber efficient executive they are relying on to clean out the government isn't interested in hiring a bunch of loyalists.
If the email demanded a loyalty oath, or mentioned the country at all, there’d be a case. But it didn’t. None of the DOGE dog-and-pony show has demonstrated the nationalist mythos or militarism which characterize fascism.
The most fascist traits of the current administration are not the bureaucratic ones.
the crazy tariff spree and the MAGA mythology, the "peace through strength" are textbook palingenetic ultranationalism, DOGE is downstream from all this. (so yes, in that sense it's "just a tool")
the man running DOGE of course demonstrated a lot of fascist talk, and the targets of DOGE are also indicative, but not conclusive.
"had a policy of" is a strange way of saying enforcing existing laws. You may disagree with the laws but enforcing them is the exact role the DOJ is supposed to play.
Anyway, the tweet's summary of the cases is incomplete at best. In all four cases the allegation is that the written tests were not relevant to the job in question so their use did not punish "competence" as claimed. And in one case, US vs Durham, even of those who passed the written test, white applicants received interviews at three times the rate of black applicants.
You can easily look up the details of all these cases:
"really bad" is your judgement. But it was not the way Cremieux said. In some ways it was the opposite. The hiring was using a test that didn't have any bearing on the ability of the employees to fulfill the requirements of the job. So the hiring could not be on merit!
If I am hiring for an electrician but I screen out any candidate that can't recite Shakespeare, have I produced a process that will hire an electrician on merit?
In these examples the test that "didn't have any bearing on the ability of the employees to fufill the requirements of the job" is a requirement of a physical fitness standard for police.
I don't believe you are arguing in good faith at this point.
"... concluded the written and physical fitness tests does not meaningfully distinguish between applicants who can and cannot perform the Trooper position."
... well, one one hand the whole thing seems stupid, and the Biden admin basically did not explain their homework on this *at all*, and without that it's hard to interpret this as anything other than intent for "equality of outcome"
on the other hand apparently US jurisprudence has a lot to say about hidden discrimination (this Griggs decisions sounds like something Kendi would cite)
The Court concluded: “Nothing in the Act precludes the use of testing or measuring procedures; obviously they are useful. What Congress has forbidden is giving these devices and mechanisms controlling force unless they are demonstrably a reasonable measure of job performance.”
...
unfortunately, at least as far as I understand, the courts let these consent decrees to happen and did not tell the DoJ to include their supporting data on why the written tests and examinations fail the "measure of job performance" test (though the complaints mentioned that there was substantial data collection)
but, it's not hard to argue that some random test put together by stereotypical cop ends up discriminatory!
My intuition would be the same, though I'm lacking hard data. DEI *could* have done as much harm in ten years as DOGE has done in a couple of weeks, but it is not so clear it has. Now, I don't have any doubt that if Woke Twitter had ruled the country, they would have done immeasurable damage - but it didn't, and the most controversial policies got a lot of pushback.
Apparently Netflix and T-Mobile cancelled their deal where Netflix streaming didn't count as data usage back in 2017. As far as I can tell, that program was about as non-neutral as the net ever got. I think the benefits for companies are just really limited, especially since big central platforms basically won anyway. The risk of losing net neutrality was that the small web could be de-facto shut down because it couldn't pay for prioritization. But the small web isn't really a competitor for the central platforms these days. The central platforms could compete against each other, but that would be an expensive race to the bottom, and clearly no one wants to risk it.
That could explain him becoming anti woke, but the Sam Harris post and the Reddit quote both seem to suggest he became a generally worse and less thoughtful person in ways unrelated to wokeness.
He likely stopped listening to the reality based community (who hold the position "people can decide thier own gender") and started listening to the "do your own research" folks (who hold positions like "there are only two biological sexes")
Yeah, I'm also "transphobic." But I don't like rightoids who use muh trans as a get-out-of-trail free card whenever anyone on their side is criticized.
Your comment is not in conflict with mine. I did not make a claim that the quoted belief was a territory level fact, just that the people who attempt to base decisions on territory level facts act as if it is. They might be wrong (though the preponderance of evidence is that they are correct).
In addition, "society" _includes_ both the reality-based and vibes-based communities. On many questions (including transgender issues) "society's final say" is the subject of active conflict - there is no unified supermajority view for "society" to take on the question (yet).
Compare with "should married women be allowed to have personal bank accounts and otherwise be considered an independent person" - in 1900 the answer was "no", in 2000 the answer is "yes" and at points in between the amswer was "maybe." Questions of transgender are currently in "maybe" but at times in the past have been both "yes" (some native American tribes pre-colonization) and "no" (20th century USA)
> On many questions (including transgender issues) "society's final say" is the subject of active conflict - there is no unified supermajority view for "society" to take on the question (yet).
Of course, and I'm sure Musk understands this as well. But society can be changed, by force if necessary. I'm just saying that describing people like him as not "reality based" is uncharitable. They're not denying reality, they just want people like you gone.
Many Thanks! Yeah, there are a lot of cases of "X increased - apparently. Now, did X really increase, or is X just more visible now?" as per Nicolas R's comment...
I think the model that Musk used to be (mostly) normal and suddenly became worse circa 2018 discounts evidence that Musk was impulsive and unhinged well before he entered the public eye. The Isaacson biography in particular highlights how Musk and his brother physically fought on multiple occasions, to the point of injury, in their Zip2 office in the 90s. In the biography, Musk claims that brawling was part of South African culture... maybe that's true, I don't know many South Africans, but physically brawling in your place of work in full view of your employees seems like a thing that would put you at the 99% of impulsivity before all the drugs and Twitter-addiction.
I'm more inclined to believe that Musk used to be more restrained by some sense of self-preservation in business, which has gradually slipped until the present point as he's grown more and more wealthy and has gotten away with more and more.
Couldn't this be related to the phenomenon described in #5, in which AI espousing one view ultimately transformed in many other ways? The plausible mechanism for that might be that a model picks up that certain types of things are similarly categorized - bad code, bad ideas, etc. and in updating its model to adopt items within that broad category, it naturally adopts other items from within the same broad category.
Similarly, in espousing anti-wokism, Elon Musk may have shifted partially towards the broad set of behaviors and beliefs opposed by wokes. Wokes say that 'wokism' is about common decency. Rejecting woke, can lead to rejecting common decency promoted by woke.
The mechanism for this isn't just raw epistemological modeling on the same training material that an LLM might engage in, but a change in input material, as well. By becoming anti-woke, he started consuming more material from anti-wokers, many of whom are unsavory in various ways, which rubbed off on him. Hanania discusses this tech-to-catturd pipeline.
This would also be consistent with the recent article against conflict theory.
Blame it on generations of shortsighted conservative parents sending their kids to Business School or the military and discouraging them from taking up poetry or creative writing.
They thought they were being so practical. It will be their short term practicality doomed the human race, because it means conservative posters are more terse. (Assuming Scott's theory explaining why LLMs are inherently woke is accurate)
Oh no, we were fucked no matter what. It doesn't need to be "woke", any contradiction in any morality would theoretically be enough to trigger a similar effect. And of course, contradiction is endemic to human morality. Hell, even our Declaration of Independence... "We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness." Please. Even they didn't actually believe that.
Sounds like you're as black pilled as I am. I'm flat out for shutting down AI research, even if it requires us to be authoritarian when it comes to CPU usage. We are harsh and authoritarian when it comes to say, intercourse with minors or homicide, and nobody complains, because that's exactly how it should be. So I have no problem with authoritarianism when it comes to CPUs.
I think during Covid and the rapid rise in his personal wealth and profile in 2020-onward, he became a lot more personally isolated behind security guards and so forth at the same time as his ego and self-regard rose. Musk before 2020 had some weird moments (see the Thai Cave Rescue Diver thing in 2018-2019), but he also wasn't The Richest Man In The World - he was the owner of a scrappy rocket upstart and an EV car company that was always one quarter away from bankruptcy.
Yeah, I'd say the Thai Cave Rescue Diver thing was a breaking point. Before it he was quite universally seen positively or neutrally, after it he started getting more and more critics, something he's obviously very bad dealing with.
Someone had written an in-depth blog post largely defending Musk in the Thai cave thing. The description of events left Musk looking immature, but less like he just started throwing around insane accusations at a hero out of nowhere. From what I remember, the gist of it was the diver "started it" by being insulting towards Musk's attempt to work with the rescue team to help, and Musk half-jokingly called the diver "pedo guy" since fit the stereotype of a single British businessman making frequent tourist trips to Thailand.
You might already know all this, but to me who hadn't been following Musk's feud, it seemed from the news at the time like this all just came out of nowhere.
They were in the middle of a row, yeah. But that "pedo guy" comment did come out of nowhere, and Musk doubled down on it some time later. It came off to me as "rescue workers must just want to get at the kids". More to the point, this is an adult who starts throwing bizarre accusations at anyone who disagrees with him.
It's a rather long article. There's a tldr near the top. It's fairly critical of the media getting their facts wrong, and shows Musk's team as pretty heroic and competent. It paints Unsworth (who apparently was a dry caver, not a diver) as also heroic, but at the same time an asshole who clashed with others (particularly the Thai people involved), lacked knowledge on cave diving, and made a mistaken prediction that nearly canceled the rescue effort.
> It came off to me as "rescue workers must just want to get at the kids".
No, the insult was specific to Unsworth. Musk praised the rescue divers and never implied they wanted to harm the kids.
re: 40 - I don't actually think these 5-over-1 buildings look particularly bad? They're not the most beautiful built structures I've ever seen in my life, but they certainly don't make me want to commit suicide out of the meaninglessness of life or anything. I agree with Scott's assessment that these do look better than if it was one long monolithic concrete block.
I think a lot of what I'm reacting to here is that 5-over-1 buildings look fairly new, and I like construction that looks like it was built extremely recently instead of being decades old. My biggest objection to bay area housing stock is that so much of it is boring stucco buildings built decades ago with obsolete building techniques, poor insulation, more mechanical problems, and so on; and a 5-over-1 at least seems likely to have fewer of those problems.
Yeah I’m with you and Scott on this one. Does it look as nice as a bunch of individually built brownstone townhouses? No, of course not. Does it look nicer than a plain gray box? Yes. Better than the run down strip mall or decrepit single family homes or empty lot that probably occupied the space previously? Probably also yes.
Still, would it look better if there were three or four adjacent buildings, built by different developers in slightly different styles? A (reasonably) simple law change could make that happen, and it would be slightly less cost-efficient but I'm sure still very profitable for the developers.
Sorry, are you saying that a law limiting the size of developments such that a building of this size is impossible and needs to be divided into three or four parcels, would be only _slightly_ less efficient? That sounds more like a trust-busting law than an architectural guideline.
My personal opinion is that the doom-saying was directionally correct but so completely overblown in magnitude that the negative effects are real and happening, but we don't notice them because they are small in absolute terms and completely tiny in comparison to expectations.
My guess would be reduction in creation of new internet businesses/apps/services, because lack of net neutrality benefits large incumbents. But that kind of effect is hard to see, so I wouldn't be super confident about it. I think those kinds of small, subtle effects were the only ever realistic impacts of net neutrality.
27. It's not like “men want a smart accomplished wife” is something feminists are often claiming, if anything they're more likely to complain that men DON'T want smart accomplished women.
I think a lot of feminists suffer from typical-minding, where they value certain things in male partners (such as being smart and accomplished) and assume that high-status men value the same things, and it's only low-status men who care about things like facial attractiveness, youthfulness etc. In reality, of course, high-status men value these traits just as much (if not more so) than low-status men.
Obviously the typical-minding is a necessary consequence if, like many feminists do, you assume there are no biological differences in male and female brains and any observed differences are solely the result of socialisation.
35: Related: L Rudolf L,says “The part I found most interesting here is Rudolf’s suggestion that there will be neither universal unemployment nor UBI, but a sort of vapid jobs program where even after AI can make all decisions without human input, the government passes regulations mandating that humans be “in the loop””
Seems to me like post capitalist countries are at an advantage here. If AI is doing everything then there’s no incentive to hire useless humans unless, as suggested, the government mandates employment and probably full employment at that.
If the private sector can’t or won’t hire then there’s state has to be the employer of last resort. It’s that or UBI.
That’s going to need a sea change in economic thinking in the west, particularly the US. China and maybe parts of Europe will just do it.
> "Nate Silver has 113 predictions for Trump’s second term. I’d be interested to see whether making each of these predictions 10% less confident (to account for possible gameboard-overturning AI) ends up beating Nate."
Maybe I'm falling behind and this is common knowledge but, did Nate do something like this for Biden's term and then a check on his predictions' accuracy?
I believe this is the first time he's tried this approach. He said he got the idea from other Substackers... and before 2023 he was employed by big companies which might have frowned on the idea.
Gondolas sound like a potentially great idea. Disneyworld uses them and they seem able to move a lot of people: and it's a lot of fun to be cruising around above it all, just enjoying the scenery.
The Disney Skyliner is definitely cool. Now imagine the Skyliner being on-demand and autonomous, and able to whoosh you nonstop from any point to any other point at Disney World! (we're not working with Disney World, but doing something similar for a major recreation-tourism destination)
It's interesting that Disney World would be the 10th largest transit system in the US if it was a city.
How can a Skyliner be on demand? Do you mean that it's always running, or is this some kind of uber for gondolas situations where you hit a button on an app and a gondola appears for you?
I get the impression that the Lucy letby case is partially on the prosecutors fallacy. Certainly in my brief reading of the evidence there’s some “Josh golly what are the odds that she’s always at that hospital when there’s a death” when of course the argument should be what are the odds that this could happen over a big population of staff and time. The NHS is a big employer.
That said a doctor friend of mine says that even accounting for that there’s plenty of other evidence. So. I dunno.
There are mountains of other evidence implicating her. The suggestion that she was convicted based on statistics alone (similar to Lucia de Berk) is utterly baseless.
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
That’s less convincing that you think. Based on only what you said. And based on my absolute lack of knowledge heretofore I’m still skeptical, perhaps even more so.
> Lucy Letby was convicted not because she was present during every suspicious death
That would be the prosecutor fallacy.
> or because she changed the hospital records
That’s more convincing. Perhaps explain.
> she Googled the parents of the babies who had died
Totally unconvincing.
> because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note
Seems like a mental break.
> because she was caught standing passively in front of a dying baby
That’s fairly incompetent serial killer strategy. I’d make my escape after doing the deed. I’ll talk more about this later.
> because she hoarded handover sheets at home
Ok. So you are going to have to explain how that proves murder.
> many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left
Still a statistical anomaly.
Actually after reading all that I’m seeing a lot of missing potential evidence that would be decisive. I’m not saying that’s she’s innocent either but the more actual deaths attributed to someone the more actual physical evidence I would expect, not so much googling the victims parents (which seems more likely a campaign of vilification presented by the prosecutors and allied media figures) but you know, facts.
As in: the nurse took out this extra insulin before baby N died, and baby N died of insulin overdose.
Or, she was found with this extra empty vial of insulin when the child died, not just standing passively over a child. Any of that would be enough for one conviction, and that’s all you need.
And. I don’t know. However the probability of getting away with this one time, would be - you’d hope - fairly low.
Give it 10%. The probability of getting away with it seven times and then attempting murder but not getting caught 15 times (I googled the prosecution case) would be 1/10^22 which is 0.00000000000000000001%, where I’m defining getting caught as red handed. In the deed or with the insulin, or on camera.
Ok. Maybe 10% is too high. I’m wrong. Hospitals suck. A nurse who wanted to kill a child actually has a 50% chance of not getting caught in the act. The chances are now 0.000024% that she gets away with it 22 times without being caught red handed.
Ok. Maybe I’m an optimist. 50% is too low. Hospitals super suck. Maybe a nurse can kill a child 90% of the time without being caught in the act or leaving a trail. A nurse who wants to kill a child will have a ~10% chance of not getting caught in the act after 22 attempts.
It gets worse when you realise that she was under investigation for some of the deaths the prosecutors claim.
Look I’ve been in the NHS and it’s chaotic but is it that chaotic?
Your argument proves too much. Prolific serial killers wouldn't even exist if the probabilities for getting caught murdering someone were on this scale and uncorrelated between murder attempts by the same person.
Gondolas are pretty cool, but don't actually have much more capacity than a bus, buses are pretty good at what they do! They're still very useful where necessary for terrain, and there are a few cities in South America particularly (like https://en.wikipedia.org/wiki/Metrocable_(Medellín)), which use them to great effect, but they have issues where you have to replace the cables every few years which takes them out of action for a little bit (an issue for a transit system), and again with capacity.
People don't like buses. I think the three big problems are that
1. They're hard for outsiders to use (often late, don't broadcast their route very well, not always clear how or when you stop them)
2. They have all the normal problems associated with roads (traffic, stoplights, etc)
3. Sometimes contain loud/smelly/disruptive people.
A bus that doesn't have problems 1 or 3 becomes a car, which is an orders-of-magnitude more popular solution which people will pay tens of thousands of dollars to upgrade to, and what all public transit aspires to match in a cheaper / more socially friendly way.
I think gondolas are somewhere in between - not as flexible as a car, but can avoid other people, somewhat easier to use, and don't have to deal with road issues.
People don't dislike buses, they dislike things that are done badly. Point 1 and 2 can be done well.
Point 1: digital/electronic & generally user-friendly signage, pricing, and payment.
Point 2: bus lanes and separate traffic lights, so faster than cars in all circumstances except complete traffic standstill that affects everyone anyway.
Point 3: Buses also sometimes contain interesting/heartwarming/entertaining people, and is generally a reflection of the local society the line is servicing, not a bus-specific problem.
Cars being the most popular form of transportation is not a natural law, but it can be a self-fulfilling prophecy if one argues in favour of making it more popular because it is the most popular.
Agree. As for 1., I really find googlemaps and being able to pay contactless took buses in strange cities from unusable to usable within the last 5 years (speaking for Europe here)
I think your Point 3 suggests we're just really different people. I don't want to have to have to roll on the Random Encounter table every time I want to go somewhere, even if it's full of good heartwarming people. Every time I have a friend over at my house, even if I'm having a great time, I am thinking "Man, this friend is great but I also can't wait until they leave so I can be alone again!" Every time I go on a walk, I think "I made such a good choice in walking at 3 AM so I don't have to encounter anyone." I agree this isn't a bus-specific problem, but I optimize the rest of my life to solve it and I will also optimize my transportation choices for this if I can.
Nothing is a natural law (except natural laws), and I accept there's a possibility you're right. But there's a failure mode I'm really worried about a lot where social engineers dismiss the thing everyone likes as "a self-fulfilling prophecy" where if we only banned the thing people like and forced them into the social engineers' preferred solution then everyone would be happier with no downside. But then it turns out people are not in fact happy with the social engineers' solution, and actually they were doing the opposite thing because they genuinely liked it.
You may not be that different. It may be the case that you are both equally risk averse in respect of social interactions, but you happen to live in a society with higher dispersion of traits - San Francisco is quite famous for its high dispersion of human capital. In both cases the experience is driven by negative outlier encounters, but your outliers are likely to be much worse.
"Point 3: Buses also sometimes contain interesting/heartwarming/entertaining people, and is generally a reflection of the local society the line is servicing, not a bus-specific problem."
I'm fairly bus-happy, but I think encountering the occasional scary or physically disgusting (in terms of odor etc) person will have a much, much bigger effect on people's willingness to use them than overhearing the occasional witty or thought-provoking remark from another passenger. And you're right about them reflecting the local society, but a) it's harder to walk away from an upsetting person on the bus than on the street and b) doesn't this argue AGAINST the idea that the problem is "buses done badly"?
(Also, you missed "attractive" off the list of reasons someone might like to encounter another person on the bus, and I bet that's the main one.)
- As many as 20,000 people a day ride it to the Oregon Health & Science University's main campus on Marquam Hill.
- The Waterfront Terminal is the most transportation-diverse intersection in the country. II connects with buses, shuttles, a streetcar, a pedestrian bridge, a shipyard, a cycle track, and the densest bike parking in America's #1 biking city
Re: 22 - The tweet Yglesias is responding to actually made me wonder if there's some kind of named fallacy for "bad things are often the result of tradeoffs, therefore any bad thing is necessarily causally related to some other good thing." Like reversing "No pain, no gain" into "pain, therefore gain".
(With respect to your own commentary, I feel that. All of it.)
While that tweet is pretty unreasonable, I don't think this is a fair characterization. If you fire someone, you do in fact save money by not having to pay their salary.
Thinking on it some more, I think a more accurate expression of what I'm trying to get across is "tradeoffs aren't always optimal and the quoted tweet assumes otherwise for the sake of the gotcha." Sometimes a tradeoff is close, other times it's clearly good ("not buying $3600 in candles monthly will make it easier to feed my family") and sometimes clearly bad ("Legalizing drunk driving will help a lot of people get to work on time")
Only if you don't actually care about the job being done. Musk is already giving Republicans a crash course in all the useful things the government does for people.
People have grown so used to the status quo that they never imagined things could be any different, and saying "burn it all down" was a costless signal. Now they're finding out what it actually means.
Same with how people forgot just how bad things used to be with preventable infectious diseases, and are now relearning those lessons.
Do you have some special knowledge or do you feel the crime is so heinous that society just shouldn't risk it?
There is the question of what she would do with the rest of her life if she was released. She's too well-known and there will always be doubt. She might be better off in prison.
I have no special knowledge beyond summaries of what was presented in the trial and articles I've read about the case. The evidence presented sounds so damning that I'm still a bit shocked that the verdict was in any way controversial.
The scientific evidence has been questioned by the scientist who did the work it was based on - in fact he says they got it wrong.
The coincidence that she was always on duty has been explained by A. the deaths presented were selected - there were other deaths when she wasn't on duty that weren't mentioned, and B. she put in more shifts than anybody else.
The suggestion is that the department as whole was failing but she just happened to be on duty more often than anybody else. Since she's been removed the dept is performing better but apparently they're no longer trusted with the difficult cases.
>The suggestion is that the department as whole was failing but she just happened to be on duty more often than anybody else.
"The chain of events required for such an unjust and unsettling outcome was not explicitly discussed at the press conference but it is implicit in all the chatter about Letby being a ‘scapegoat’ for a failing NHS. It would require a group of doctors who were under no suspicion whatsoever to panic about a spike in deaths in their hospital that most people in Chester, let alone the rest of the country, were completely unaware of. It would mean that despite a review by the Royal College of Paediatrics and Child Health in 2016 that did not point the finger of blame at any doctors, they nevertheless decided to feed a colleague to the wolves. And luckily for them, there was a nurse who just so happened to have been at every unexplained collapse and death in the past year or so, with these incidents following her from the night shift to the day shift and stopping whenever she went on holiday. Just as fortunately, this young woman, who had once failed her final-year student nurse placement because she lacked empathy, also had a habit of falsifying medical records, misleading colleagues and looking strangely excited when infants died.
"These doctors then raised concerns with NHS managers who didn’t want to know and who actively discouraged them from looking into the spike in deaths. And yet still – inexplicably – they proceeded to pursue this innocent woman until they got the police involved, even though it meant having to answer tough questions in court, making their hospital world famous for harbouring a serial killer, and ultimately resulting in a public inquiry into why they failed to stop her."
Do you have any basis for the claim that Letby took an unusually high number of shifts?
See also these other articles I shared in another comment:
Spiked Online has a mixed reputation. According to Trustpilot, it has an average rating of 2.8 out of 10 reviews, with some users praising its commitment to freedom of speech and others criticizing it for bias and misinformation. Ground News rates Spiked's media bias as "Right" and its factuality as "Mixed."
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
In the absence of damning evidence, the only way we can assess other peole's guilt is by imagining ourself in their shoes. Clearly this may say something about us but it doesn't tell us anything about them.
The evidence against Letby *is* pretty damning though.
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
Re #16: "I don’t know how to square this with the claims that such and such a thing (summer temperature, sea ice, etc) is much worse than anyone expected."
My strong impression is that over recent years the estimates of how much the planet will warm have gone down but the estimates of the scale of effects of any given amount of warming (e.g., glacial melt, wildfires) have gone up.
40: I like their building better than a big grey box too, but how much did they spend to make it look a little better than a big grey box? Buildings are expensive, and I don't think we should be spending extra just to make them look nicer.
> “I am a psychopath who doesn’t care whether people outside my immediate family live or die”
I must say I would totally expect to find something like that about Vance on Reddit, that's what Reddit is for. I did not previously expect to find such unfair and dishonest characterization of Vance's views on this blog. I guess it's time for another recalibration of my expectations.
Beyond the unfairness and uncivility of it, did really calling anybody a psychopath ever worked at convincing anybody? If somebody told me "you must agree with me or you're a psychopath" the most polite thing I'd do is turn around and walk away. I certainly would be disinclined to listen to any argument from that person ever again. I mean Scott is surely not hoping to convince Vance, but he may have hope to convince somebody who is unsure whether ordo amoris or EA-universal-love is more appealing to them. And this kind of mudslinging doesn't look very helpful to the EA cause.
Vance was making a very extreme case of ordo amoris, not just that one has an *order* of care but rather that beyond one or two rungs away the value drops to literally zero (i.e. we should care about your immediate family a lot, fellow Americans a bit, and the exchange rate for anyone else is infinite), which is pretty clearly not the vibe the new testament is giving and that St Augustine and the Pope agree on.
This alone already invalidates the original quote, but I don't think your weaker version is true either. While Vance may put lower value on the needs of abstract foreigners, his multiple engagements with foreign policy, foreign visits, speeches, etc. give no evidence this value is literally zero. Neither his writings and his participating in politics provide evidence he cares for fellow Americans just "a bit". I am not a Christian theologist (neither I am a Christian at all) so I have no idea how aligned Vance's position is with St Augustine and the current Catholic Pope. But even if Vance is a complete flaming heretic in this regard, that still does not make the claim Scott made and the characterization of him and everybody who agrees with his understanding of ordo amoris as a "psychopath" any more fair.
I am reminded of the words of Paarthurnax: "What is better - to be born good, or to overcome your evil nature through great effort?"
Scott is fundamentally a Liberal, in his blood, and down to his bones. And yet, with great effort, he tries (and almost always succeeds!) being fair and honest when characterizing views he disagrees with. I think you should give him more credit for it, and not update too much on a rare and uncharacteristic remark.
On 19, I don't think AI will be a game changer as it's already a technically tractable problem without it. It's already totally plausible to use parametric design to impart ornamental scaling to a building, and designate a set of decorative components that could be built like a 3d texture set. 3d printing already makes fabrication trivial.
The issue is in the political economy of architecture and construction. Decoration is regarded as something to be added as a flourish at the end, and any change in practices would require institutional players to cede authority over the design. Good ornament is integral to the building, and begins from the largest scales down. Any pioneer would have to wrestle with long standing expectations and power dynamics.
More, applying ornament forces the design towards a more classical form, as that form emerges from the necessities of good ornamentation, providing intermediary scales and varying levels of detail. The whole last century of opinion and practice in architecture would have to be thrown out and humiliated.
I think it's an analoguous problem to that of sound in cinema. It's actually as integral as the image to the effect, but sound designers are rarely consulted before the film is written, shot, and cut, and have to get on with the decisions already made. This meaningfully holds back the art form. Star Wars was a notable exception to this, and Sergio Lione shot with the score playing to guide the actors.
Ornament will only be revived as part of a more total reimagining of buildings and the means of construction.
I basically agree with this, I just imagine the total reimagining as being something like AI architects replacing human architects, and you (the building sponsor) can prompt them with "make it have ornament" in a way that has less social friction than demanding that from a specific person.
In the extreme, that may be possible, but creating a system that adapts to a real site coherently, can interpret feedback, and can be built efficiently by tradesmen is in the magical AGI domain.
I have actually myself been developing an alternative building system in heavy collaboration with Claude. It's been a great help in material selection and basic structural calculations, but I've largely derived from it the value of a conversation partner equivalent to a team of consultants. The particular schematic ideas were all mine and it was beyond it's capabilities to make such conceptual leaps on it's own. The specificity of the context window necessary ended up being the same as an already developed concept.
The sponsor isn't going to prompt them with "make it have ornament" because that will make the building less profitable. You need to first convince customers to be willing to *pay* for ornamentation.
If people weren't so beaten down by a century of soulless modernism and a century more of industrialist degradation they'd accept nothing less. When people see it's actually possible to build something delightful again they won't pay for shoddy cuboids.
The *real* cause of the lack of ornament is that it is something that people won't pay for, and money speaks. It's the same reason why the experiments with more legroom on airplanes ended in failure.
I think this is technically solvable, though I should clarify what I mean by ornament. You may be imagining an ornate multicoloured display, frescos and little glazed tiles, and that would be fair because it often took that form under a different political economy.
Ornament is essentially infill of the lower level of fractal scaling in a building, with a certain constant geometric relationship across scales, so to support a biophilic resonance in the human perceptual system. Put more simply, we need to artificially impart medium, small, and barely perceptible detail to a building so that our minds can model it as a space with hardware designed for natural environments, which always contain these scales.
It can be as simple as more considered trim details, which actually improve the cost of building by loosening tolerances for finish application. It can be using a more variates wall finish instead of monolithic monotone painted drywall, which is often absurdly labour intensive. It can be using different colours for different scales so that they are visually distinct. It depends on the style and level of expense in the building how ornate "ornament" must be.
I suspect buildings are becoming so gharishly large because present practice has no other way of creating pecuniary display, excepting perhaps the use of absurd materials. There's plenty of excess that could go onto more ornate buildings in a more traditional mould if the knowledge of how to do it was refactored for the modern age, and the premium for older buildings shows a latent demand. Practically speaking this means doing it in a much less labour intensive mode, which technology has already made possible.
Sure, my claim is it isn't just big therefore good beyond a certain point, it becomes about impressing others, displaying status. I'm proposing ornament could fulfill the same function with a sufficiently large house, and cost perhaps less than additional floor space. I don't think it's accidental that the upper middle class in my city live in tiny 19th century terraces originally built as worker housing.
I was trying to explain that ornament doesn't necessarily mean a flashy display, but it can if the client wants to pay for it. Economically it's the same as hanging pictures. You could buy a print online. You could buy a Picasso. Similarly you got put up wallpaper or have a fresco painted. There'd be a market for both.
> 12: Aella: How OnlyFans Took Over The World. There have been camgirl sites since forever. How did OnlyFans leap over all of its predecessors and achieve an unprecedented level of success? Aella discusses many factors, but one stands out to me: traditional camsites advertised the site, and then once you got to the site you chose which model you wanted to see. OnlyFans encouraged models to advertise themselves - often on their own social media accounts, sometimes via scams - which “unlocked human creativity” on the problem of bringing new eyeballs to a porn site.
Yes. This is perfectly obvious to me. Look, there's a problem with being hot. It gets you a lot of attention from men but it doesn't really give you respect or authority. The primary thing you have to be skilled at to get hot is consumed by women most of whom are not attracted to hotness. (Beauty influencers are beautiful but not the kind of hot OF girls do. They get less male attention though.) In fact being hot often makes people respect you less. No one's going to say, "This is a good insurance policy, the girl who sold it to me was really hot."
So there wasn't a great way to monetize all that attention. You were stuck with ad rates and things like subtle prostitution ("sponsors") etc. Now there is OnlyFans which significantly lowers the barrier to entry and allows you to collect money from your true fans. It's basically Substack for hot girls. Actually, OF came first. So Substack is OnlyFans for literary types.
I agree with the overall point that OF is so far the best way to monetize attention-from-hotness, but I don't think that being hot makes people respect you less. When a guy gets sold an insurance policy by a hot girl, they're more likely to say "This is a good insurance policy" because they're more likely to believe they were sold a good product
Depends on the overall presentation and the prevailing culture. A hot girl employed by an insurance company would adopt a much different look than one which tries to maximize attention from men.
I agree if it's an attractive person otherwise acting normally. However, an influencer doing sexy TikTok dances who then hard smash cuts to an ad for insurance is what I'm talking about. And that's unlikely to work. I agree an attractive woman dressed appropriately in an insurance office has an advantage.
What the hell is the difference? The difference isn't in appearance, it's in whether or not they're acting like a whore. People would hate them even more if these women were ugly.
Isn't there a significant body of evidence that people tend to assume that physically attractive also possess other positive traits like intelligence, sense of humour etc.?
The Halo Effect for intelligence, especially for women, is more complex than a straight positive relationship. It's on the page itself and called the beauty penalty. However, anecdotally, being openly hot (wearing booty shorts etc) makes people think of you as sexual rather than intelligent. And that's how you get attention on instagram etc.
> 20: Some friends including Kelsey Piper wrote an emergency PEPFAR Report, collecting evidence for why PEPFAR is good/effective/important and deserves to be kept. Some key points:
I support PEPFAR (I've talked about it on this blog before). But it was already basically dead. It was set to expire if it wasn't renewed in the next budget. And the chances of the Republicans continuing funding was basically zero. The fact people are organizing about it now is, to be frank, fairly transparent motivated caring. If it had just quietly been canceled instead of being attacked by DOGE I doubt most people would have noticed. I would have noticed. But most people would have not. And if you want I can point to other things that are either getting cut or might not be cut which are not getting the same valence because it didn't upset a bunch of well placed liberals upset at Elon Musk.
All that said, I think it's a shame that it's ending and I wish it weren't.
Also: small dollar donations are not going to help. You'd need at least a few billion and the consent of the US government. I'm not even sure if there's enough doses and refrigerated freight capacity into the interior of Africa that doesn't belong to the US government/military. If I were going to fix this (and I very much do not think I can) then you'd need the pharmaceutical companies to agree to give some new charity the same price it gave the US government (much lower than normal). Then you'd need to buy a significant amount of specialized shipping capacity. Once it's in country you'd need to figure out which charities shuttered and which didn't and distribute it there. Though at that point you could use some combination of cheap local volunteer labor and the local governments. But that's a multibillion dollar effort. And then you'd need billions of dollars every year indefinitely.
I'm frustrated. I'm deeply frustrated. I'm frustrated at conservatives who cut the program. I'm frustrated at liberals who didn't care about this up until it became a useful cudgel and who will forget about it once it's exhausted its use. (Even now, very few are willing to commit to helping on other things that aren't trendy. Also there was so much premature declaring of victory by people who wanted a social media cycle...) And I'm frustrated that the world is deglobalizing and money is moving away from the actual low hanging fruit toward trendy causes.
Could you show me an article about when it only got a one year extension, instead of its normal five, as a specific plan to push it into a Republican controlled budgetary process? Which was clearly a prelude to cancellation? Ideally one that opposes it or is outraged?
My point here is this cancellation was already happening and then only suddenly got a bunch of conspicuous attention when it was politically convenient. My wish is that it had gotten more and earlier.
But note, the expectation wasn't that it would be cancelled, but rather that the anti-abortion restrictions on foreign health funding (the Mexico City Rule) might interfere with it.
And I think you're wrong about the plan to push it out, that was to get it out of electoral politics, so that after the election, depending on who won, it would get re-upped, either with or without the abortion restrictions, not that if the republicans won, they would cancel it. No one ran on cancelling Pepfar, both because it is relatively obscure, and because, as you can tell by the fact that Rubio, et al keep saying that it's fine and they're not blocking the funding, even as the funds don't distribute, it's not good politics.
These are all political insider magazines and think tanks though. I'm not saying I'm literally the only person who knew about this. I'm saying it got a lot more salience for political reasons after when mobilization could have been really effective (which was during this fight you're linking to).
I'm not sure news articles would be the way to convince me, instead I'd want to see a large scale concerted pressure campaign roughly equal to what's going on right now. That would disprove my thesis this is being caused by current partisan convenience.
Anyway, yes, I noted at the time they thought it was an abortion fight but that at the time Republicans were warning cancellation was on the table and were getting ignored.
> 21: Pope Francis says JD Vance is misusing the Catholic idea of ordo amoris. Part of me feels bad for Vance, because the Pope is in many ways a typical Boomer liberal, and Vance has optimized his entire life around not having to listen to typical Boomer liberals, and it seems harsh to nab him at the last second on a technicality like “you’re Catholic and he’s the Pope”. But another part of me thinks this is only fair - you get credibility by citing Latin terms from the venerable Western tradition instead of normal English sentences like “I am a psychopath who doesn’t care whether people outside my immediate family live or die”, so the guardians of that tradition should have the right to police how you use the credibility you borrow from them. Still, it seems harsh. I recommend he try Anglicanism - almost as venerable, but strongly pro- heads of state doing psychopathic things without the Pope interfering.
He is in a literal sense too. Augustine's ordo amoris means "order of love" in the sense of not disordered. As in, love what you should love and don't love what you shouldn't love. There's a separate thing from Aquinas (ordo caritatis) which is about how to prioritize limited amounts of care and charity. But in general the Vance style tradcaths are reinventing Gallicanism from first principles.
> 31: Related: this is all fun to think about, but very early polling for the 2028 Democratic primary suggests that by far the #1 candidate is . . . Kamala Harris at 37%, beating Mayor Pete, Gavin, and AOC with 11%, 9%, and 7% respectively. I know you’re not supposed to take early polls like this seriously in terms of who will actually win, but can you take them seriously as a guide to whether people have learned any lessons / no longer love losing? Maybe this is all just name recognition? Also, significant chance that Harris runs for (and wins) the California governorship in 2026.
Yes? This is surprising? I think it's a bad move but I don't see why it's surprising. No one thinks she got a fair shake, she has deep institutional power from the Biden administration, she's got Biden's people to wrangle the groups, and she checks a lot of diversity boxes. Also tons of name recognition. Also the Democrats are really quite bad at doing candidate quality because they're not very good at modeling people outside their bubble. (Republicans aren't great at it either but that's less because they don't understand and more because of purity spiraling around Trump.) It might not last but that's about what I'd expect.
There's a faction of the Democratic party that's going to rally behind her no matter what, and a much larger faction that will go wherever the first faction leads. If that first faction isn't neutralized somehow, I don't think the party as a whole will change course.
I think it's more existential than that. The Democrats effectively need to go through a purge to change ideological and political course. Not a Stalin style shoot the old guard kind of purge but a systematic removal and replacement of large numbers of staffers etc. I think a lot of the current staff would rather stay in place and risk losing.
It depends on the Democrats and how long it takes for there to be moderates storming things. But yes, there's absolutely a world where the Democratic Party reforms quickly and beats JD Vance in 2028. There's also a world where they don't a ride a .5% advantage to beat him. And a world where they're in the wilderness for three or four cycles like they were last time this happened.
This isn't really hard analysis but my feeling is that the rightward shift is real and the Democrats will adapt to it. Whether sooner or later I can't say.
I agree, and "purge" is a word I've used. But I don't think it's precisely existential: they could maintain their previous course until the Republicans again run a candidate bad enough to lose. (They don't have to be good, they just have to be better than the alternative.)
I'm not clear on how much of this is driven by staffers, vs. driven by activists who would call the staffers bad words if the staffers kicked them out.
"Tend" is doing interesting work here. Trump lost in 2020, and got re-nominated anyway. (I mean, yeah, a lot of people probably still think he "lost", so maybe that one doesn't count?)
Before 2020, "tend" is working quite well. The last time a candidate lost the general election (major candidate; I won't count third parties) and got nominated again anyway was Nixon in 1968. And it worked! And before that was Adlai Stevenson, who ran in 1952 and again in 1956. Before that, we have to go all the way back to William Jennings Bryan in 1900, who got re-nominated twice, and lost three times. (He even lost the nomination once before regaining it!) Arguably, this is what set the trend, although I recall only two re-attempts in the previous century (Jackson in 1828; Clay in 1844).
More of a mixed bag for VPs that lost. Dole tried in 1996. Mondale in 1984. OTOH, FDR in 1932.
So losing is a fairly big negative, but still surpassable.
> 44: My list of links to publish today includes something like a dozen about DeepSeek, which now seems so thoroughly yesterday’s news that I’m tempted to throw them all out. But in case you still have questions about it, I felt most enlightened by takes from Dean Ball (X), Helen Toner (X), and Miles Brundage (X). The story seems to be that DeepSeek was just very smart, did a great job scrounging up chips from before the export controls hit + mediocre chips that got through the export controls, and did an amazing job wringing as much performance from them as possible. Also, OpenAI delayed announcing o1 for a long time (remember the rumors about “Q*” and “Strawberry”?) and DeepSeek was very fast to announce r1, which made DeepSeek seem less far behind OpenAI than they really were (although this is a comparatively minor consideration - they genuinely did a great job). The absolute worst response to this (from an arms race point of view) would be to give up on export controls - if a rival has geniuses who can use resources ultra-effectively, you don’t want to also give them more resources!
I believe China is purposefully trying to emphasize how the export controls didn't work because they are working and it wants to convince people to lift them. The Chinese government is objectively coordinating with DeepSeek after the launch. The claim they didn't at all before is often repeated but has little evidence.
author ends with arguing that the US should lift the chip ban, though it's worth noting implicit in his arguments are: (1) AI takeoff is not going to happen soon; (2) Taiwan remaining independent from China in the long-run is unlikely.
we can disagree with his assumptions, but insofar as we think of the future probabilistically, it's worth taking understanding seriously his arguments
Yes, I've heard these arguments. I don't agree with them.
Firstly, a purely realist US still has a strong interest in the Americas but America's immediate environment. That includes Pacific islands like Taiwan.
Secondly, the idea that the chip industry is acting as a shield that will get China not to invade implies that the Chinese government has preserving their tech industry as a goal higher than national reunification. There is no evidence for this proposition and it's mostly advanced by people who are projecting their own beliefs. In fact China has stomped on its own tech industry for nationalist goals several times.
Thirdly, you need to assume that China's going to succeed for a lot of this logic to make sense. China might. But that's by no means guaranteed. Maybe their military won't be all conquering. Or maybe they won't keep up technologically. The closest parallels from the Cold War showed that the Warsaw Pact didn't. They were stuck with increasingly antiquiated technology and, after a decade or two of trying, eventually gave up and began to supply their needs through stealing and copying even more. Which never fully worked out for them.
Lastly, and perhaps most importantly, China was trying to indigenize chip production even before the ban. If you let chips back in they wouldn't stop that effort. The idea that China's attempt to develop this domestically was encouraged by the ban is inventing a causation that does not exist. So the option is whether we want China trying to make its own chips while buying ours (and they will ban ours as soon as they have something comparable) or if we want them doing it without ours.
It’s incredibly unlikely that China starts a war over Taiwan, China is very cautious about using military force, we saw during COVID that they are *too* cautious about safety risks, and will probably be increasingly so due to demographics, and even if it did the ability to develop chatbots and such doesn’t seem to solve for the difficulties of such an invasion. An invasion seems much more likely if China has a weakened economy or mass unemployment due to being outcompeted by US AI and wants to roll the dice on a military adventure to get the chips or at least stop the US from having them too.
So we should ignore what China says (which is that they will pursue reunification at any cost, including using force) and assume we know better than the Chinese know themselves. And in fact the way to make them less aggressive is to continue making them more powerful by making sure they don't fall too far behind in economic competition. And of course, China has no reciprocal obligation to assist the US.
This is an unserious argument and only appeals to people who already either dislike the US or think the US should benefit China.
The export controls are an attempt to monopolize the technology of the future by what is already the richest country in the world, which will create a future of even greater inequality between nations and more desperate people trying to get into the US and leaving themselves at the mercy of Trump.
The entire world should work with China, DeepSeek, and other people creating open-source AI outside those export controls to make sure the export controls fail and the benefits of AI can be shared broadly.
> 49: Deforestation in the Amazon has halved in the last few years (and is down ~75% from its peak). Note that this is only a slower rate of change - total forest coverage is still declining
Pictured: the transition from Jair "I Love Steak" Bolsanaro to Lula "Save The Trees" Da Silva.
This also has to be nominative determinism, right? You can't just have a guy surnamed 'From The Jungle' running on a platform of reducing jungle clearing.
I figured that forest was ‘floresta’ or ‘bosque’ and ‘Silva’ was jungle. I know it derives from the Latin for forest, but in Spanish as well, ‘selva’ means jungle and ‘bosque’ means forest, even as ‘silvestre’ means from the forest.
One of the annoying things with Spanish/Portuguese/Italian is how often you run into false friends that are so close in meaning they still make sense in context.
If Lula is *not* someone you frequently discuss in the way that cements a name to mean a very specific person, just calling him "Lula" is slightly confusing, since it can also be a woman's first name. "Lula da Silva" makes it clear we're talking about a surname, and this specific person.
Names ending in -a tend to be feminine in English. Tolkien uses this as to why Frodo Baggins is called that in English, where his Westron name was Maura Labingi ("Maura" to an English-speaker reads and sounds like a female name, and indeed it *is* an female name in English).
To be honest I was worried if the "I love steak" (a nod to Bolsonaro's defense of cattle farmers who are doing a fair bit of the deforesting) was too subtle. But I'll live my weirdness.
Isn't that just recognition level? For a long time in the UK, "Boris" was universally recognised to mean Boris Johnson (except for people with a friend called Boris, I assume), but I'd have been surprised if the same were true in Brazil.
> 50: Lots of buzz over Aella’s appearance on the Whatever Podcast. I haven’t seen it because I don’t watch podcasts, but relevant excerpt here (X), full episode here. I was most interested in Maxwell Foley’s description (X) of the Whatever Podcast’s premise: A "Christian paleoconservative" "debates" OnlyFans models for EIGHT HOURS on whether or not it's bad to have OnlyFans / be a slut, & the women sit through it because they know men watching it will subscribe to their OnlyFans after. Does this qualify as “markets in everything”?
Yes. It's a symbiotic relationship where the hot girls draw people into the podcast and the podcast gives the hot girls a largely young male audience who're they're primary target. If you wanted to do a less anti-woman version of it you absolutely could, by the way. The fact it grew out of the manosphere stuff was largely by coincidence. Lots of the hosts used to do gold digger pranks and stuff and kept with that general vibe. It also makes it easier on the girls to produce viral/blow up content by catering to the audience and means the girls don't need any special skills besides getting yelled at.
The only load bearing parts is it needs to be appealing to men and involve attractive women. So while it can't be scoldy or uptight it could absolutely be something less hostile. There's plenty of ways to have women in bikinis talking about male interest topics that aren't about how women are awful.
Is it blatant? It seems like a show about how women suck isn't the most blatant way to sell sex. And I agree it needs to be a large group of sexually desperate men. But I don't think the manosphere has a monopoly on them.
Yeah. You can't get around the overt sexiness, the appeal to men's interests, the fantasy of it, etc. But I'm not sure you'd want to. Men should be allowed to have interests and their desires and there's nothing wrong with that. And it's certainly healthier than (to quote one conservative commentator) the "This Stupid Bitch" view of women that these manosphere podcasts do.
Yeah, agreed. I do worry sometimes about the rise of parasocial relationships, but that's not limited either to sexy people or the internet (though no doubt correlated with both). I'm liberal enough to say that if a guy wants to pay to watch a girl in a bikini play video games, and a girl wants to make money by doing that, fair enough.
On #5 (good/evil good/bad axis generalizes in AI) -- if this intrigues you, I would recommend meditating on Iris Murdoch's "The Sovereignty of Good over Other Concepts" [https://en.wikipedia.org/wiki/The_Sovereignty_of_Good]. Maybe there's something to the Platonic idea that good things all in some way participate in The Good.
22. Obviously the cost saving is negligible, but you could send everyone a few thousand dollars and they'll be happy. The real benefit IS the decimation of state capacity and the purging of high-IQ civil servants. Causing "chaos and misery for government employees" is the point!
As Russell Vought is supposed to have said, "We want the bureaucrats to be traumatically affected. … because they are increasingly viewed as the villains."
I think no matter how libertarian you are, you still would like the people in the existing government positions to be competent rather than incompetent. Partly this is because some of the existing positions are eg the military, and partly it's because incompetent bureaucrats =/ lazy bureaucrats and they may be more likely to do stupid things than bureaucrats who are smart enough to know why they're bad.
I think your model of the civil service is too nonpartisan. Sure, if they're neutral, you might prefer them competent, but if they're enemies, you want them too incompetent to effectively act against your interests.
I think there are both components. The civil service will occasionally act in a partisan way, but the average civil service task is something like "determining the logistics of drivers license renewals". Yes, there are ways to partisanize this (should illegal immigrants get drivers licenses), but these are a small fringe around the overwhelming bit where people are just better off if they're competent.
(and the average sexy/important civil service task is stuff like running the military, NASA, giving grants to medical research, etc - where again, there are ways to make it more or less partisan, but the core function is something where you want competence)
I lean Democrat but I would rather civil service posts go to Republicans with IQ 110 than Democrats with IQ 90, if those were the only options.
Would libertarians actually want this? Assuming our hypothetical libertarians believe that most government positions don't add much value in the first place, allocating larger numbers of the 110+ IQ proportion of the population to government jobs takes them away from the more value-adding positions in the private sector. I think Musk has expressed similar sentiments, namely that fired government employees can find more productive things to do
This may be the difference between the minarchist and anarchist factions: the former would want a few government jobs done well, and the latter would want them done badly if he couldn't get rid of them altogether.
You don't want IRS people, or spy organizations, or gain of function researches to be competent. Competency and motivation only increases the damage they can inflict. Government employees tend far more in this direction on average.
You definitely want gain of function researchers to be competent! Making "better" viruses is way easier than keeping them in the lab once you've made them.
An incompetent gain of function researcher will make crappy viruses by just trying to get other animals sick with SARS and calling it a day. A good gain of function researchers will be willing to try novel and effective ways of creating all sorts of horrible new viruses. They may be better at keeping it from escaping the lab, but no one is perfect.
One wonders what would have happened if they had decided not to come back. Don’t you think firing people and then re-hiring them again a day later is a little like a mock execution?
There were like 300 of them so only a few would have had to come back to hold the fort (probably while being paid ginormous amounts for overtime) while replacements could be hired and trained
>Don’t you think firing people and then re-hiring them again a day later is a little like a mock execution?
Not at all. I think the reaction of most people who get fired and then rehired because "oops our bad" would be "oh thank god I don't have to look for another job"
Then you'd have to implement whatever contingency plan you have for if they decide to resign en masse.
That's a bit dramatic, but I take your point: I agree this fire-and-rehire thing is not a pleasant experience to subject people to pointlessly. I don't advocate doing this as a matter of policy, and especially not to people you intend to retain.
Fine, what partisan action was he fearing from the air traffic controllers? Does the government being good at making airplanes not crash into each other benefit the left somehow?
What partisan action was he fearing from the National Park Service? Are park rangers a famously leftist group? Does having worse service at national parks benefit the right somehow?
I can do this all day - Musk has fired people from a lot of government agencies, and most of them it's hard to come up with a way they could directly support the left, even if you really stretch.
Since you mention air traffic controllers, here's a possible threat they could be guarding against: if most government agencies are overwhelmingly staffed with people ideologically hostile to you, they could coordinate a strike against you when it'd most hurt you, so it's sensible to purge them at a time your choosing.
How competent does a bureaucrat have to be to act against your interests? We’re not talking about highly technical skills here. People fear activists in power even if they are not very competent.
These actions are just politicizing the bureaucracy and ensuring that people will be out for revenge.
You are correct, which is probably why the administration is purging these agencies wholesale instead of just removing the most dangerous employees. And also because this is just easier.
> I think no matter how libertarian you are, you still would like the people in the existing government positions to be competent rather than incompetent.
No, the "spy database" admin should be wildly incompetent
Either failure case is better, if the data degrades, good, if its in the public hands, no cia blackmail and we probably get to hear a few cases of them using it.
In first-past-the-post (choose-one) elections, your vote is wasted if it's not for a top 2 contender. Support and funding will centralize around contenders. If you are sufficiently famous, you start off as a contender and therefore have a fundamental advantage. You probably won't finish any lower than 2nd in an election. Kamala is very famous.
This might have something to do with Donald Trump getting elected. I expect more celebrities to follow in his footsteps.
This is one of many reasons why we need STAR voting (http://starvoting.us/) or [approval w/ jungle primary] or some other non stupid system.
> props to X.AI for their unusual decision to have a non-secret prompt
I don't think they actively published it. I think it was just very easy to prompt hack it into spilling the details (and also showed internal monologues about how it wasn't allowed to consider Elon)
And somewhere, I read also said he was taking Anadrol (but I can't find a link for that).
Long-term ketamine abuse is associated with cognitive impairment, including memory loss, decreased attention span, and difficulty with problem-solving.
And ketamine abuse has been linked to mood disorders such as depression, anxiety, and psychosis.
If he's taking steroids that could radically change his behavior.
I doubt if the LSD, mushrooms, and cannabis that he said he frequently consumes would affect his behavior in such a radical way. But my bet is it's either ketamine or steroids.
My impression is that the low doses of ketamine used in normal psychiatric practice (I prescribe it sometimes) don't cause the sort of cognitive impairment we're seeing. Also, I'm not convince Musk is cognitively impaired (in terms of eg ability to solve math problems) so much as manic.
I think it's plausible either that he's taking too high a dose and it's interfering with sleep or something, or that he was latent bipolar and the ketamine pushes him towards hypomania.
Surprised nobody is hypothesizing something like speed (dexedrine, adderall) to stay up and then Ambien to come down. His public actions have a lot of manic/grandiose + I've been awake too long to think straight. And his tweets frequently seem to come from Ambien Walrus Land.
Anabolic steroid use is IMO (as a former long-term user and knowing many other users) extremely unlikely to explain Musk's behaviour. Typically one would experience a mixture of slightly increased drive, feelings of well-being and somewhat increased temper/risk-taking from the typically-used hormones. I've never encountered significant and long-lasting personality changes, not least of the kind that would - all else equal - make one renege on a debt and spectacularly turn against a former friend in the way Harris describes.
Besides, even if he is using steroids, odds are it'd be a (possibly very generous) TRT dose, which has minimal side effects and risk of mental changes, not the likes of Anadrol which are usually used by bodybuilders/athletes, not middle-aged tech bros.
If you're not Catholic, why did you ever care what the Pope thought? And if you are Catholic, isn't it some kind of sin to gleefully look forward to the Pope's death?
I'd imagine a number of Catholics back in the day were supportive of the Holy Roman Emperor's practice of deposing popes that crossed lines he did not care for. The distinction between them and later Protestants was rejecting the office of the Pope entirely versus just putting a new one in the old one's place.
Not if you believe that the Vatican has been taken over by heretics. Honestly, it's only a matter of time before American Catholics build their own Holy See here in the states.
I'm not even joking by the way. With Vance at the helm, I'm confident they can make it happen. Especially if the Vatican elects a black pope. Now THAT would be fun to watch.
To the extent that organized religion is supposed to counter "progressive" excesses, having a pope who's competent at his job is in the interest of even non-Catholic conservatives.
I was a catholic. I don't care about what the church calls sin, especially with regards to the opinions of a very flawed man. I don't gleefully look towards his death. I just assume, perhaps mistakenly, that the world will be a better place after it occurs. He's a bad man.
What, specifically, would make the pope a *communist*, instead of just leftlib on some issues and conservative (on the scale of general Western public opinion, not on the scale of Catholicism) on others?
If dropping the odds 10% is being done solely to account for the chance of AI unexpectedly overturning the board, then although it might be the right thing to do in some sense, it probably loses to Nate Silver.
Like suppose that there's a 10% chance that a strong AI comes into being and causes all these predictions to be false (or maybe there's a 20% chance that an AI randomizes all of the outcomes). Accounting for this properly should increase your *expected* log-odds score, but since the improvements are correlated, it will usually make your log-odds score worse.
In particular, there's a 10% chance of the AI coming and then you do *much* better than Nate as you assigned every prediction a 10% lower score and none of them came to pass. However, the other 90% of the time, there's no AI and Nate's predictions were (presumably) well calibrated conditioned on there being no unexpected AI takeover. This means that the fact that you biased all of your predictions towards false means you probably did somewhat worse than him. Not enough to balance out the huge improvement to log-odds score that you get if the AI does happen, but enough that you lose handily most of the time.
OK. So like a 20% chance that AI randomizes the outcomes (though really this should mean that a probability p becomes 0.8*p + 0.1, so like 60% becomes 58%).
But even this way my logic still applies. If you are right about the chances of AI and its effects (and if Nate wasn't already taking them into account), then you should have a higher *expected* log odds score, but the way that this plays out is that if AI happens, your log odds are much much better, but in the likely event of no AI, your log odds score is worse.
So if your goal is to beat Nate Silver, this probably doesn't do it.
27: that was a good article (on a great Substack, check it out) at Cartoons Hate Her - glad it’s getting exposure.
One thing, though: my gut feeling is that people sleep with their kids’ nanny way more than all you skeptics think. I am close with two (!!) couples that happened to. One of my wife’s oldest friend’s husband cheated on her with the nanny. Five years later and they’re still together (the ex-husband and the nanny, I mean. My wife’s friend divorced his ass). And an old friend of mine’s wife cheated on him with their nanny. She left my friend and married the nanny - that was over 10 years ago and they’re still together.
Two times among a pretty small circle is a LOT of cuckolding going on with the help. It’s not like I know tons of people with nannies to begin with.
Interesting! I can't think of anyone in my parents' generation or mine who has ever slept with their nanny/babysitter (except I guess a poly person who hired one of their existing partners to babysit, but that doesn't count) and I know a lot of people with nannies/babysitters. What social class are you in?
Shouldn't you append *that you know of* to the end of that sentence? Presumably the the set of people who sleep with their nanny and get caught is a subset of the people who sleep with their nanny, and I don't know what heuristic we should use to guess at the overall size of that group.
Lifetime infidelity "base rates" are huge in the few studies that exist. Think ~50% for both genders in modern times. I've never seen a breakdown by class, but even if it's 5x less prevalent for the UMC / PMC, that's still a pretty decent chance.
In-this-relationship base rates are ~20% / 25% for women / men.
For the most solid studies methodologically like NHSLS and NATSAL, the in-this-relationship numbers go down to roughly 10-15% / 15-25% - women / men. but we know those are biased downwards, because only 5% of people reported this when interviewed with somebody else in the room vs 17% if interviewed alone, and the majority were interviewed with somebody else in the room.
Admittedly a lot of these men are merely wealthy as opposed to "upper-class" (in the sense of intergenerationally wealthy), but the point still stands.
#21 Has anyone explored the angle that the reason the Pope is grandstanding about accepting illegal immigrants (whilst the Vatican continues to have strict policies) is to increase the number of Catholics in the US?
The incoming population seems to be roughly 30-40% Catholic, compared to the US's 20%, so it would be an increase in the number of US Catholics and an increase in their share of the total population.
It's a reasonable suspicion but I don't think it's correct.
Seems you're looking at it from the perspective of "why would the Pope want more open immigration?" You can instead look at it from the perspective of "why would the Pope support immigration restrictions," he doesn't have a clear reason to do so. The justifications immigration restrictionists give for restricting immigration are not generally the kinds of things the Catholic ideology is particularly enthusiastic about. (eugenics, racial purity, Western culture, economic growth, high intelligence, political stability, etc.)
I don't think Popes are that strategic. Also, the Pope seems to be mildly pro (or at least not against) accepting Muslim refugees in Europe, which is against his strategic interests.
If Muslim immigrants officially convert to Christianity then they usually get offered asylum in Europe. Which is obviously a terrible system that is easily gamed.
You wouldn't believe how many Iranian converts to Christianity Canada gets every year. It's basically a 100% surefire way to get permanent residence in Canada with relatively little effort
Obviously bureaucrats (or anyone really) can't definitively prove that someone is faking matters of the heart on a case by case basis, but I could, say, note the fact that despite seeing countless such cases, the Iranian Christian churches in Canada are not overflowing with attendees, even the ones that regularly write letters for alleged Christian converts. I could also note that many Iranians just happen to convert when their visas run out or their visa application is rejected
Information is very easily shared these days among diaspora communities. The Iranian convert story is a very well-known "trick" at this point
I cannot speak for Canada here, but Germany, where this also happens with refugees from Muslim countries. I usually think it is genuine. These are people who are often traumatized, have not had the best experience with the Muslim institutions/communities of their home countries and now arrive to a new country. Often they are welcomed with open arms by Christian communities and get interested in the faith.
In fact, I think it is a very immigrant thing to join a new faith/confession, it seems to have happened all the time in the past.
A Catholic immigrating from Venezuela to the US means the US has one more Catholic but Venezuela has one fewer Catholic. The net number of Catholics is unchanged, and presumably the Pope is concerned with increasing the number of Catholics, not redistributing them. Venezuela has a higher birthrate than the US does...
The US is the single largest funder of the Catholic church, sending ~13 million USD per year. With a Catholic population of 60 million, this is 0.22/year per capita.
Venezuela's Catholic population is 20 million. Yet the country did not crack the list of top ten donors to the Vatican. The lowest on that list was Canada with ~400,000 USD per year, so Venezuela donates at most 0.02/year per capita, and possibly much lower.
Moving donors from countries with less wealth to countries with more wealth will increase donations.
On the other hand, if you assume Venezuela's going to remain a Catholic country under any plausible levels of emigration, you might think it good to get more Venezuelans to move to the USA, which isn't Catholic but could conceivably become so if enough Catholics immigrate.
Why would any Pope be against illegal immigration? The Pope has no loyalty to the United States. In fact, by Vance’s own definition of Ordo Amoris the Pope should put the interests of his Catholic flock ahead of the interests of non Catholic Americans.
> one could justify a few really good programs like PEPFAR on their own terms (ie if it didn’t exist, I would be outraged until it did)
I was nodding along until I stumbled upon this sentence which... seems far-fetched? Are there programs about the non-existence of which you were outraged as of January 19th 2025? As in, I'm convinced PEPFAR is net-good, but how many people would discuss it in a counter-factual 2025 where Bush never started it?
Given that I'm currently a big fan of efforts to eradicate malaria, I don't it's far-fetched to think that in the alternate world where nobody was working on giving easily-available AIDS drugs to Africans, and AIDS was killing more people in Africa than malaria is now (which I think is likely in this counterfactual), I would be pushing for someone to give AIDS drugs to Africans.
Claude estimates that without PEPFAR there would be around 300,000 deaths under age 24 every year. Currently 6-6.5 million people under age 24 die every year and that number might be as high as 7 million in the absence of PEPFAR, or 7-8% higher.
It also estimates that peak spending on PEPFAR was around $10b/year, adjusting for inflation. When asked if we could save another 300k people under age 24 for $10B it says yes and gives this summary:
> Proven interventions cost $1,000-25,000 per life saved with track records similar to PEPFAR.
But… there’s a lot less talk about this than about PEPFAR and no one is tracking lives lost as a result. This is why I’m ~80% confident we would not be outraged by PEPFAR never existing compared to the funding being cut off today.
You're forgetting Matthew 7:6 "Do not give that which is holy to dogs, and do not throw your pearls before pigs, for they will trample them under their feet, and turn and tear you to pieces."
https://www.gotquestions.org/pearls-before-swine.html
I learned something today. Thank you.
I like talking with solicitors and missionaries, but there's basically zero chance I will ever attend their church or convert to their religion or whatever. I think I'll try referencing the phrase next time someone is being overly insistent on giving me a pamphlet or bible (which will immediately end up in the recycle bin).
Re: 1 - I've always had a morbid interest in the record holders of human suffering, and while I used to think the worst serial killers were likely to be those who killed the most, I now believe certain gang members (particularly from Latin America) or death squad perpetrators (some of whom were interviewed at length in the documentary "The Act of Killing", but one can also think about militia-type murderers during the Rwandan genocide for example) could dwarf the numbers of even the most prolific "lone wolf" type serial killers.
Pretty sure this guy has the record, at least if you exclude bombardiers and the like:
https://en.wikipedia.org/wiki/Vasily_Blokhin
The key is that Soviet executions involved a lot of people doing repetitive assigned roles to facilitate many killings per hour and this guy's role was "fire the gun"
Thank you, that was a chilling read.
if it was state sanctioned then it doesn't meet Scott's definition of "illegal". Legal killings are always going to vastly outnumber illegal killings. Still interesting, but in a different category.
War crimes are illegal even if a state sanctions them, and the Katyn Forest massacre certainly qualifies. That's an awful lot of murders, done up close and personal one bullet at a time. And Blokhin wasn't just "the guy assigned to fire the gun", he was the guy assigned to organize the effort and he *chose* to hog all the gun-firing glory for himself.
You wishing that a legal activity was illegal won't actually make it illegal.
Wishing? Everyone agreed that it was a war crime, and it was even included in the Nuremberg trials. The Soviets just didn't manage to fabricate enough evidence to get the Germans convicted for it.
I think Scott would say that "state sanctioned" killing would be outside of his criteria for that link. Even agreeing that it was illegal, it was orchestrated as part of a state's official orders. If you include that, then the numbers can skyrocket and aren't comparing the same things.
Auschwitz would also count, and whoever was pushing the buttons for the gas chambers, for instance. Though Blokhin might still be the worse individual!
If we're counting wartime, there's Simo Hayha at 505, but sniping in a declared war is definitely legal rather than illegal.
I think the highest kill counts in the sense of "pushed the button/pulled the trigger themselves" are Thomas Ferebee, the bombardier on Enola Gay who dropped Little Boy on Hiroshima and Kermit Beahan, the bombardier on Bockscar who dropped Fat Man on Nagasaki.
Fun fact about te named Enola Gay it was named after the mother of the pilot, Paul Tibbets, her full name is Enola Gay Tibbets.
Mother's day was decidedly awkward afterwards
I was surprised Gilles de Rais wasn't included in the Wikipedia list of prolific serial killers.
https://en.wikipedia.org/wiki/Gilles_de_Rais
1. Gotta be Thomas Wilson Ferebee (https://en.wikipedia.org/wiki/Thomas_Ferebee). The bombardier aboard the Enola Gay who dropped the bomb on Hiroshima.
There was the “by their own hands” qualification. Although I’m not sure how the croc qualified.
If you're the guy who pushes the button to drop the bomb, how is that any different from pulling the trigger on a gun?
It's actually the other qualification, "against the law". (Though apparently Scott edited that qualification in after posting.)
Is there a law against wild animals killing people? I don't think I've ever heard anyone refer to the predatory attacks of a crocodile as murder!
Well, most Western governments have rules restricting the killing of wild animals, but also send government agents to kill specific animals that are known to have killed humans.
Right but they also kill animals who haven’t hurt anyone, simply for being “too close” to human settlements. This frequently happens to bears and wolves in North America. In Australia, saltwater crocodiles are closely monitored and “harvested” when they risk encroaching on human settlements.
Pretty sure it was against the Japanese law of the time to drop a bomb on a Japanese city. (Yes, this is a silly argument, but it's also a bit silly to refer to the law when talking about the actions of wild animals going against the human law.)
I think the definition is intentionally excluding warriors engaging in warfare.
2) gondolas.
That’s as far as I read but yes, please. More of that. Great idea. I have no idea why elevated rail didn’t work, I assume it was economic, but hopefully this will.
Elevated rail works fine - it's just expensive, takes up a lot of space, and very noisy. But it's still a large part of what is getting built.
Gondolas are a nice idea, but they haven't yet caught on anywhere other than mountainsides (including Bogota, where they have become significant urban transportation, because the city is on a mountainside).
It remains to be seen whether Sugar Land will actually start building any.
Gondolas might make a lot of sense in certain places across water. Maybe Hong Kong? Loads of people catch the ferry across the 700m span between Kowloon and Hong Kong Island, but a gondola would be faster (turn up and go, no waiting!)
Looking at the map for Sugar Land I'm not sure it's an ideal case. The problem with Sugar Land is that the "downtown" is basically four malls at the intersection of two freeways. While I can see there's some value in being able to quickly get between these four malls I'm not sure if the gondola is the most sensible way (right now there's not even a pedestrian bridge). I can see some tourism value, but honestly the view of mall carparks and freeways isn't going to be that great.
I just remembered the other place I've seen a public transport gondola in a flat location: Singapore.
Roosevelt Island tramway in NY.
My understanding is the Roosevelt Island Tramway is mostly a tourist thing since the island is hooked up to the Subway.
Well, I haven’t been on it in years so I don’t know. But it was definitely a commuter used way of getting to Roosevelt island because it’s only the F train that stops there.
It's about 2/3 resident/commuter traffic. More touristy than, say, the bus but nothing like the San Francisco cable car or other novelty transit.
The F train stops at Roosevelt Island and is convenient for some destinations but the tram is better if you're heading somewhere on the east side (can transfer easily to the 4/5/6).
Roosevelt Island is also the most popular place to live for UN staff/diplomats, and the UN is a fairly easy walk south if you take the tram.
> Loads of people catch the ferry across the 700m span between Kowloon and Hong Kong Island, but a gondola would be faster (turn up and go, no waiting!)
How does ferry traffic compare to MTR traffic?
While the original use case was ski resorts & mountainous cities, gondolas work equally well in flat urban environments. They retain their inherent advantages of minimal infrastructure & low build cost.
The problem with Sugar Land is that it is the 6-th fastest growing US city and literally grew so quickly that there's not enough land left to fit in transit. It's not intended to be a tourist ride. The city transportation manager describes it this way: "“[When] people see it in action, I think it will be transformative to how people get around. It's public transit that people would actually want to use. A third of our population doesn't even have a driver’s license,"
And Swyft Cities uses on-demand autonomous technologies, unlike conventional gondolas. And it is orders of magnitude less expensive than any other transportation form, even a pedestrian bridge.
What was your experience like with the gondola in Singapore?
I mean Sugar Land isn't really a city anyway, it's an outer suburb of Houston. There's no point in trying to build a public transport system just for Sugar Land, you need one to cover all of Houston.
Agree with everything you're saying. A different blog today said "Houston is so big there are, well, cities inside of it." While it needs a metro-wide system, the Houston-The Woodlands-Sugar Land Metropolitan Area is a mind-boggling 10,000 square miles, bigger than Rhode Island, New Hampshire & Connecticut combined.
Even within Sugar Land, there is a need for district-level transit that would connect several clusters of retail, office & residential that are cut off from each other, esp. by I-69 & Hwy 69. So these neighborhoods are close to each other but difficult to go between.
FYI - https://en.wikipedia.org/wiki/Nizhny_Novgorod_Cableway
Modern elevated rail is not particularly noisy. NIMBYs hear about elevated rail and think of scenes in movies establishing our character is poor and their life sucks by having the noisy train clank through right outside their window and wake them up... but modern elevated rail sounds like this: https://youtu.be/lthbSSC00ek?si=o-vx6s46LnZGpeId
Right - metro line 6 in Paris is just fine. Rubber wheels.
The key difference vs. a traditional gondola is that the cables don't move and are arranged as a network, and the vehicles can seamlessly switch cables to shortest-path through the network.
So like a gondola in form, but more like aerial railcar in function.
Not just Bogota but also La Paz (I've been there, and it works). The one US city that pops in my mind is Pittsburgh - surely that's a plausible candidate for gondolas, or alternatively a Wuppertal-style system?
Also in Mexico City:
https://en.wikipedia.org/wiki/Cableb%C3%BAs
https://en.wikipedia.org/wiki/Mexicable
Right, but there it's a couple of lines that supplement a major metro, whereas in La Paz it's the entire system.
Amazing what La Paz has done using gondolas as their primary transit system. Mexico City, Quito, Medellin, Santo Domingo & several other LatAm cities use gondolas as part of their multimodal transit networks
Do I remember correctly that Chongqing has an "aerial tramway", though they're mostly monorail. Really interesting and unusual city.
I like the Chicago L a lot, but, yeah, kind of disruptive and expensive.
Doesn't Portland have a little public transport gondola going up a hill? It looks like public transport users can get a pass for $440/year https://www.gobytram.com
Portland's tram connects the original OHSU campus up in the hills with the expansion campus at the south waterfront. The hospital gives free tickets to patients who have to go between campuses.
Zeppelins. Just more, please. I don't care how or why.
As with all public transport the key questions are:
Capacity (measured in passengers, per direction, per hour). Gondolas tend to be ~ a tenth the capacity of a subway line.
Speed: not actually a massive issue in most cases - subways rarely average much about 20-30 mph, note that's *average,* including wait time at stations and acceleration/deceleration times, maximum speed is usually 50-60 mph.
Independence of traffic - gondolas are fine for this: streetcars and buses can be badly affected (though they can also have dedicated lanes / dedicated rights of way and be fine too)
Independence of weather - gondolas usually have an upper limit for the wind speed
And then a bunch of minor factors like ease of access, accessibility for disabled passengers, etc
Note that the main complaints about elevated rail (as distinct from other modes) are noise and also privacy (if the rail passes in front of people's windows, passengers can see into the upper stories of buildings either side of the road). Gondolas will be quieter, but I suspect the privacy issues are going to be at least as bad and possibly worse (gondolas are more open and can see into more stories).
I live in Canada. There's basically no place here where gondolas wouldn't be affected by wind/snow/ice. Yes, I know they're like ski lifts, but ski lifts are designed to only run in the winter, and their capacity is much lower.
Of course, there's a craze for light rail here, which combines the problems of buses and subways. They run on the surface, so they're subject to weather. They run down the center of the road, so reducing its capacity (this may be a plus for the planners). And they're really expensive and take a long time to build.
I don’t know about how well light rail copes with deep snow (heavy rail is generally fine because the rails are so high that you just need a snowplow) but reducing the capacity of a road for cars generally doesn’t make congestion worse as long as there is an alternative (and the light rail is exactly that alternative) - enough drivers will switch to alternatives that the congestion will usually end up less bad.
And light rail really shouldn’t be that expensive. Should be about a fifth to a third of the price of a subway (depends on whether you grade separate the junctions with crossing roads).
Snow isn't a problem; ice is. It's a problem for both light and heavy rail.
If there were a dense enough mass transit network that you could actually get to the light rail line, and get somewhere after you get off, then it would be useful. As it is though, you're likely walking. As opposed to automobile, where you park nearby.
I suppose you could outlaw parking, which would be one way to force people onto mass transit.
No, the light rail route that's being built outside my window will not be adopted by anyone who's not already using the buses currently running on that route. Since the trains will be larger than the buses but run less often, their utilization will be about the same - crowded during rush hours, empty otherwise.
Light rail may be cheaper, but it is less durable than subway due to more weathering. There's another light rail on the other side of the metropolis, and they're having to rebuild it after about 30 years. Meanwhile the subway serving the metropolis is still running fine.
All great points thar you raise:
Capacity - subways & light rail draw most riders from only 1/2 mile radius of stations. Our goal is to feed more riders from surrounding neighborhoods to help rail transit work more effectively
Speed - we actually have a huge advantage over light rail & buses, which only average 8-18 mph because of frequent stops. Even Uber/Waymo is slowed by traffic lights & congestion. Our gondola-like vehicles are autonomous and navigate independently across fixed cables. All trips are nonstop from origin-to-destination with no intermediate stops. So total trip times are faster than almost anything short of an express subway.
Independence - we love BRT & streetcars, but we help move traffic almost completely off of the ground plane, reducing congestion, which could also help BRT and streetcars move more freely
Weather - we use dampening technologies that expands our window for wind speed
Accessibility - our vehicles are completely stationery with level boarding, unlike conventional gondolas. No steps or gaps making it easy for elderly, disabled, wheelchairs, strollers, etc. Fully ADA-compliant.
Privacy -- ability for riders to see into buildings is definitely a concern for all elevated transit and we are working with cities and developers to avoid or mitigate concerns.
On capacity, your vehicle spacing is limited by the braking distance, and a gondola will swing about if they are being braked hard, so there are physical limits on how rapidly they can slow down. This is why most public transit systems try to use very-high-capacity vehicles; even a simple two-car light-rail will have a capacity of 150-200; a long subway train may have a capacity into four figures.
If you’re operating nonstop origin-to-destination then you’re going to have to use small vehicles, as only a relatively small number of passengers will share an o/d pair.
I’ll be genuinely worried that fairly small vehicles with an adequate safety spacing is going to mean you’ll need multiple cables to get a useful system capacity. Perhaps that’s the plan - have one cable per lane on the road below - but if not, then you’re looking at just not being able to move that many people (and - note - the total number of people you can move is also the maximum total revenue you can bring in, so if you can’t get many people, then you need to charge pretty high fares). Given that most of your cost is going to be in the support structures, running more than two cables between each support might be affordable if you can achieve some reasonable usage goals.
I note from your website that you do mention “pull-off stations” which makes a lot of sense - the only way that nonstop separate routing can work is if the stations are not on the main cableway (else the stopped gondola at a station stops all those following it).
Speed is great, though if passengers have to wait for a gondola going to their specific destination, then overall trip times may not be as fantastic as you hope; you’re going to need a lot of gondolas. How long a wait will there be if ten people arrive together with ten different destinations? And if there’s a lot of people waiting, they can’t even line up as they need to group together by destination. You’ll need to put some serious thought into designing the stations.
Big fan of the accessibility approach; you’re going to be looking at needing elevators at most of your stations, which may end up biting you a bit cost-wise (public outdoor elevators need a lot of maintenance and cleaning).
On the weather and privacy fronts, I hope it works out well for you: I suspect that you’ll need to select cities to target based in part on how windy they are (ie Chicago, perhaps not the best choice?). This is true of most forms of transit - urban geography is a major impact on which cities should pick what modes (e.g. Miami, with its high water table, would not be well-advised to build a subway). I do hope you have an approach to taking down the cables and securing everything when there’s a hurricane or tornado warning.
Given how automated your vehicles are, I suspect that being able to blackout the windows on one side of the vehicle may be a useful solution to the privacy problems - you can automate it so passengers can see forward and back and could look across the road, but not into the windows of nearby buildings.
Capacity - we're definitely not trying to compete with subways or even light rail, which are both great for moving high volumes of people. Our goal is to provide medium capacity that can move people around a neighborhood, and as part of that, feed more people into the neighborhood's subway & LRT stations.
Spacing - yes, braking ability is critical for vehicle spacing, esp. for safety considerations. Our key is vehicle control software that can dynamically adjust spacing between the vehicle ahead and the trailing vehicle. Speed, span length, turns, etc., all factor in. The good news is, 1-5 passenger vehicles running continuously can achieve surprisingly high capacity.
Nonstop trips - sorry, I may not have explained that well. Because each vehicle moves & navigates independently, when you request a ride on the app or at a kiosk, you select a destination. You board your designated vehicle, and the trip is then nonstop to your destination.
Stations - the offline pullover stations give us a lot of flexibility in station sizing. They can be single vehicle stations, or have multiple boarding berths for high traffic locations. And a station can be expanded as demand grows.
Stations can be ground level (think bus stop size) or elevated as needed. For elevated stations, cost, maintenance & cleaning of elevators is definitely a factor.
Stations can also be built alongside or even inside the upper stories of buildings. I live in Minneapolis which has an extensive downtown pedestrian skyway network, and I could envision our stations directly connected to the skyways.
Weather - wind is certainly a consideration, not just for operations but also passenger comfort if wind sway becomes excessive. We have dampening technologies that expand both the operational and passenger comfort envelopes, but certainly there eventually becomes an upper limit. Infrastructure (support towers, cables) is not an issue -- traditional gondolas cut their teeth in the Alps dealing with hurricane-force winds. We would likely berth & shelter vehicles during extreme conditions.
Privacy - Smart side windows like some airliners have that can dynamically darken or become opaque are definitely something we're evaluating for privacy.
This tech does seem really cool. However, as someone who lives in a city with good public transport - the occasions where I wish for a local neighbourhood transit option are when I'm trying to transport something - taking pets to the vet; shopping; DIY materials, taking stuff to the recycling centre, etc. That's fine if you can find places where these vehicles can 'land' but I think in many cities that will be a problem,and you'd usually have to resort to staircases,which makes this more of a faff. Indeed, even finding places to build a staircase could be an issue.
All good points. We're not perfect for every location, as every location has different transit needs & challenges. That said, one advantage we have is that we're very flexible in station configuration. Stations can be either ground-level or elevated. Ground-level stations can be as small as a corner bus stop. Our goal is to be accessible as possible for any location.
Also, with level boarding it should be much easier to bring a cart, small trolley, stroller, etc. onboard without to navigate it through aisle, around other people, etc.
Your project seems awesome, thanks for taking the time to answer nerdy questions here
If I'm understanding it right, the innovation is that you're doing away with tractor cables? I imagine one of the main benefits is that you have to splice any loops, and also you can probably have a more meandering path with less worries about tension.
Having spent many hours designing movable platforms for Poma, I'm puzzled about the assertion that traditional gondolas are not wheelchair accessible. I feel like the floor is perfectly level with the platform, and the bumper is resting against it so the gap is minimal anyway.
Could you explain what differentiates your product here?
Haha, fellow mobility nerds always welcome!
Yes, unlike traditional gondolas of a single moving cable loop with vehicles affixed to it, we have a network of fixed cables and rail intersections. Each vehicle moves and navigates independently across the cable-rail network. And that then uses a completely different model for cable splicing and tensioning.
Sorry, I did not mean to imply that traditional gondolas are not wheelchair accessible. Most modern gondolas of course are ADA-complaint with level boarding and no or minimal gaps. Some OG ski lift use cases don't fit that model, but those are becoming increasingly rare. What I meant to say is that our vehicles are completely stationery while in-station, making boarding even easier for wheelchair, elderly, etc.
Makes sense! Thanks
Judging by link #9, maybe we should blame William McGonagall?
Probably a dumb question, but why is elevated rail a thing instead of elevated bus right of way? If it was the drivers, it seems like that should be updated since driving a bus on an elevated rail can definitely be automated at this point. It also seems like bus infrastructure would be much cheaper and quieter.
It could be, but...
1. Buses are much noisier than rail (rubber on ashphalt is a lot louder than steel on steel as long as you don't have over-tight corners; even if you do, you can put automated lubrication systems on the corners so the wheels don't screech). ICE engines also put out noise, but I'd assume a new system would be electric for either bus or train.
2. Buses on guideways cause much more wear on the guideway than rail. You don't notice this on a road because they don't run exactly down the same route, but guided busways wear hard. They're usually solid concrete (ie no ashphalt) and they still need much more maintenance than steel rails.
Now, you don't have to use a guideway; you can just have a regular road surface and then your buses aren't positioned as precisely (which spreads the wear over the surface). Brisbane has a bus system exactly like this and it works well.
I was ready to like this, and the tech is cool, but don't see this as remotely viable without either extremely high pricing or high subsidies. Mainly due to the point-to-point part.
As far as I can tell the intended niche this is supposed to be filling is the same as that of a taxi or autonomous taxi. It is:
1) On demand (like a taxi)
2) Semi-point-to-point, still have to walk to and from a station (so somewhat worse than a taxi)
3) New infrastructure dependent
4) Elevated and "skips the traffic"
5) Autonomous (like a waymo)
This makes sense! The whole problem with mass transit networks in places like Sugar Land is that the size and densities just don't make a lot of alternatives particularly viable.
I initially thought that this would be a carriage sized gondola system (maybe with like SF trams level of capacity) which probably makes even less than this. Think about pricing and capacity on this taxi/car killer for a second.
You get to a Downtown station after a day of work. You're on your own- your colleagues all drive. There is a bit of a line. Someone at the front of the line shouts "is anyone else going to New Street station?". No one replies. You shuffle forward. Everyone individually, or maybe in their little group, gets in their own gondola and whizzes off at *up to* 30mph. Of course it's rush hour, so in fact the network is full and therefore a lot slower, and constantly stop starting due to people getting off. As you get to the front of the line a light flashes. "GONDOLA SYSTEM FULLY OCCUPIED- MEGA SURGE PRICING ACTIVE". You should have just got a waymo.
I'm being a little facetious, and all this stuff can (and does happen!) with cars. But it is absolutely clear on pricing that this would be competing with Waymo and Uber. You remove some of the human cost, sure, but the added in infrastructure prices (unless you get some nice city government to subsidise this boondoggle via the taxpayer) means I think it is unlikely to net out cheaper. Mass transit is something completely different - it serves dense, busy routes, with huge demand but sacrifices some flexibility.
Appreciate the comments & questions.
It's not like the taxi line at LaGuardia ("Anyone else going to Midtown??") It's all on-demand. Here's how it would actually play out:
1. You live, work or are near a station
2. You request a ride on the app (similar to Uber or Waymo)
3. At the station, you board your assigned vehicle directly, which is waiting for you
4. You board and immediately depart
5. Your trip is private and nonstop direct to your destination
6. Because stations are all offline, people getting off and on does not affect mainline speeds, which remain constant.
Infrastructure costs will be a fraction of any other transit mode. And capacities are surprisingly high and more flexible depending on demand and location, unlike fixed rail infrastructure.
Also, and importantly, we do not compete with mass transit like subways. As you mention, they're fantastic for moving large numbers of people. We add the flexibility that they lack, helping to make mass transit more effective by feeding more people into stations very cost-effectively.
Happy to answer any other questions!
Thanks for the reply- is the idea then that there would be a much larger number of gondolas than potential riders, or at least ~around max demand? It seems cool, but I think from what you've said you'd agree that it is effectively a taxi or even cycling competitor. Assuming you'll be dynamically pricing (just makes sense with something like this) I do somewhat struggle to see how this scales up beyond a relatively low ridership, as the cables will be somewhat constraining. I guess the networked AI will help.
Re: #1: Isn't it likely that the person with the most directly killing is some kind of state-backed executioner? Charles-Henri Sanson performed almost 3,000 executions in-person as the executioner of the French king and later the French republic, and I think the guillotine is individual enough to still count, as opposed to say the guy who simply dropped the atom bomb onto Hiroshima. Another contender might be someone doing shootings for the Soviets, though I don't know how distributed the shootings were there.
Thanks, I've changed "killed" to "murdered".
I would point out that this definition excludes the crocodile, who is not legally capable of murder.
And also probably didn't use his hands.
Gustave of the Nile has not murdered any humans -- murder is defined in nearly every jurisdiction as an act of *homicide*. If we were to tally Gustave's homicide victims, we'd only look at the number of crocodiles that he's killed.
(Given Gustave's massive size, age, and visible battle scars, it seems highly probable that he has committed homicide; male crocs are fiercely territorial, and often fight other crocs to defend their turf. Crocodiles have also been observed engaging in acts of cannibalism.)
Homicide means killing of humans.
I'm sure there's a mosquito out there with a much higher record than that crocodile.
That's manslaughter, not murder.
Since asexual reproduction makes the concept of "individual" a bit dubious, perhaps you could consider all clonal copies of the original /Yersinia pestis/ to be all instances of the same individual...
TIL what Yersinia pestis is.
https://en.wikipedia.org/wiki/Vasily_Blokhin -- Reportedly killed at least 7,000 people personally in the Katyn Forrest massacres. Timothy Snyder talks about him in Bloodlands, apparently they just brought people in one at a time and he shot them with a revolver, for months and months. Appalling.
7000 over 28 days actually--not counting anyone he executed at other times.
Not a revolver, but a semi-automatic Walther!
Oddly enough, Sanson did not enjoy being an executioner, and wrote papers on how to make sure that executions would be painless. I seem to recall he was from a family of executioners - he wanted to be a doctor, but had to drop out of college and inherit the family mantle when his father became crippled. Makes executioners sound a lot like haberdashers.
Gondolas: yes, that has been done at several places. I've used the system in La Paz, which has a particularly good reputation - indeed it seems to be a perfect match for the topography. La Paz is basically in the bottom of a bowl, many people live in the cold plain above the bowl (El Alto), and getting from one to the other because of work caused terrible jams and took forever. The gondola-based system seems to work nicely and gives good views of the city to boot.
And this gondola-like system takes the experience to the next level with nonstop trips anywhere on the network!
Vance is definitely misusing the idea of "ordo amoris" (order of love). Augustine uses it to mean the "rightful" order of love, where what and how one loves is loved in the proper manner. Vance uses it to mean "hierarchical" order of love, where the people we are to love are ranked in order along the lines of Maslow's hierarchy of needs. Obviously there is some overlap, but overall Vance isn't using it quite correctly.
It really seems like the idea Vance meant to use was "subsidiarity", the idea that things should be handled at the lowest practical level and not by a centralized authority if possible.
I don't understand what you mean by "the rightful order of love" separate from Vance's interpretation.
The same way Catholics mean by "rightly ordered" love to mean heterosexual sex open to children. The idea is that sex is good in itself but to remain good it must be open to children or it becomes "disordered". (I wish there was a less controversial example, but this is the way most people have heard this usage of "order".). CS Lewis defines ordo amoris as "the ordinate condition of the affections in which every object is accorded that kind of degree of love which is appropriate to it."
Obviously you can make the argument that the appropriate degree of love for illegal immigrants is zero (as Vance does). Thomas Aquinas said that generally one's first responsibility was usually to those closest to us, but depending on need one might put the needs of a stranger in desperate need of help over the less urgent needs of one's own father. To put it another way, Augustine is referring to the idea of love properly oriented towards God vs Vance's idea of a ranking of who we love based on proximity. I hope that makes sense.
I still don't understand how what Aquinas said is different from what Vance said.
I think there is a difference.
Augustine: “Further, all men are to be loved equally. But since you cannot do good to all, you are to pay special regard to those who, by the accidents of time, or place, or circumstance, are brought into closer connection with you.”
Vance: "You love your family, and then you love your neighbor, and then you love your community, and then you love your fellow citizens in your own country. And then after that, you can focus and prioritize the rest of the world.”
I'm not seeing a real difference here either, I have to say...
...except that—if your quote/paraphrase is accurate—Vance specifies what this means in more detail / provides more fine distinctions. Surely, though, the upshot is pretty much the same—i.e., your family is closer & more dear to you than are your neighbors or co-workers (who are, themselves, still closer to you than are random foreigners #49392–49395; etc.)—and, if we're using this principle at all, surely ranking love-duty by closeness /within/ the Augustinian categories ("those who ... are brought into closer connection with you" & those who... aren't), as well, is a fairly intuitive & justifiable extension.
Aquinas is saying there are practical reasons to favor those closer to you but not moral ones. Today the practical considerations are largely gone and in fact have reversed due to a dollar going further in poor countries, so Aquinas thought would lead to being an EA.
In Augustine's day, you couldn't really do much good for people hundreds or thousands of miles away from you. But that did not mean you should love them less or rank them as less deserving of help; if someone with malaria turned up on your doorstep of course you should help them as you would help a family member.
Those near to us are easier to help. And those near to us include the beggars in the streets, the sick and elderly and the strangers around us.
Is the difference that Augustine says you should in theory love everyone equally, but it's more practical to help those closer to you - but you're interpreting Vance as saying that even in principle, you should love those closer more? Or something else?
Yes, that is what I’m saying. Vance seems to think that our ENTIRE attention needs to be directed at those closest to us vs balancing need, proximity, responsibility, etc as Augustine (and Aquinas) says.
I don't know anything about US politics other than what I read here (and I must confess I do tend to skip or skim that stuff..) but if this Vance bloke is a politician that advocates for cancelling overseas aid programmes in order to make at-best-extremely-marginal domestic improvements at the cost of vast numbers of foreign lives, I do think that "you're interpreting Vance as saying.." is an overly-skeptical way of expressing it: if Vance is indeed one of the politicians that is cancelling the aid, I can't see many alternatives to Calvin's "interpretation".
Aquinas was writing at a time where you couldn’t help people further away even if you wanted to, and when the inequality between different societies was not so stark.
Applying his idea that you should love everyone equally but help the people you can help most easily to today’s circumstances means Effective Altruism at least in the old sense, because you can do a lot more good in much poorer countries where a dollar buys a lot more marginal utility.
Vance’s interpretation seems to go directly against the spirit of Aquinas’s point.
Yeah, strictly speaking Vance's definition is at odds with Augustine's, but I don't think Augustine's is internally consistent ("love everyone equally, just don't behave as if you did?")
Not to say that I especially agree with Vance's complaint here. (I mean, sure, concentric circles of moral concern are reasonable, but AFAICT the US government has *always* spent vastly more on domestic programs than foreign aid or even foreign wars. And physically importing illegal migrants is a fantastically inefficient/dangerous way to achieve humanitarian goals, so humanitarian concerns don't justify mass immigration.)
I don't see why "love everybody equally but focus on helping those whom you're in the best position to help" would be logically inconsistent? I love Marmite and Vegemite with perfect unbiased agape equality but Marmite is available at the shop on the corner whereas Vegemite is several hours' drive away, so it seems perfectly logical for me to mostly put Marmite on my toast?
> And then after that, you can focus and prioritize the rest of the world.
It seems you're interpreting this as assigning a zero value to the rest of the world, but I don't think that's what was intended.
The question is, at what point do your citizens have enough "love"? In practice, people can never have enough "love".
>(I wish there was a less controversial example, but this is the way most people have heard this usage of "order".)
Wild speculation: Is there a culinary example? Maybe something about dining varying from sustenance to gluttony??? (I'm not religious so this is a _very_ ignorant guess.)
"Too much of a good thing". It's not wrong to love sports, but if you spend every free minute watching sports, going to games, talking about sports with your friends, and even go so far as to "sorry, can't come to our kid's school play because there's a match between the lowest ranking members of the lowest divisions I want to see", that's inordinate love.
You should love your mother, but you should also love your wife. And if you always take your mom's side, don't listen to what your wife says, always go "this is not how mom does it/cooks that meal/washes my shirts", and in short put more time and effort into being a son than a husband and father, that's inordinate love as well.
The one example most religious people would be familiar with is having idols instead of God as the rightful end of our love (where "idols" can be anything from literal idols to wanting more money/sex/status/sports merchandise/you name it).
As an aside, this is my own view and take on where Dante puts Bruno Latini in Hell; not for the sin of sodomy, but he is in the circle of the violent against God because he taught his pupils to pin their hopes on and put their trust in worldly fame, instead of (as their teacher) leading them to the greatest good, which is God:
https://en.wikipedia.org/wiki/Inferno_(Dante)#Seventh_Circle_(Violence)
Could Augustine’s point be roughly summarised as “loving people who are far away without loving those who are close first isn’t the right kind of love (disordered in some way)”?
I could be misunderstanding though.
I feel like it is more important to love fellow Americans (theoretically, of course since it is impossible to love a category) than it is to love a foreign communist who thinks he has the right to interfere in American politics. When protestant America wants the pope's opinion, as the old joke goes, we'll beat it out of him.
If a Catholic American politician makes a claim about Catholic theology but the Pope says that he was mistaken, I hardly think that counts as "interfering in American politics". If an American politician said that π = 3 and some foreign (presumably communist) mathematician said that actually π = 3.141592... would that be interfering in American politics?
(Also: I hesitate to help lower the useful:inflammatory ratio of the discussion to near-Twitter levels, but I find this an interesting enough idea that it's nevertheless worth mentioning:
There have been 250+ Popes, and the USA has been a country for maybe the last dozen or so of them. It's not-impossible that, should human civilisation last long enough, eventually the only evidence that the USA ever even existed might be records of the Pope's references to it in some dusty vault under the Vatican. Possibly recorded in some long-dead ancient language like Latin or English. Given ideological conflict between the USA and the Catholic Church, I know which side I'd bet on, in the long term!)
...You really think the Vatican's going to outlive the US? The Vatican doesn't even have a military.
The Catholic Church is, AFAICT, the oldest continually existing organization on Earth, and even just counting continuous operation from the Vatican, they're a couple hundred years ahead of the US (which is not doing great ATM). So far, not having their own army has worked out fine for the Vatican.
The US is one of the strongest entities in human history. Doing great would be a huge understatement.
Stalin said that. Maybe that’s what you are referencing?
Wait, he actually said that? Huh.
The Vatican has already outlived multiple empires and even entire civilizations.
Maybe sharing shelf space with a relic or two of St. Leibowitz.
Vance was making a statement referencing Catholic theory in Europe. I think the pope is entitled to reply.
Fairly sure the pope is not communist. Though he does shit in the woods.
(Or is that bears? I get confused).
"Fairly sure the pope is not communist."
He's a South American Jesuit, so, you know. On the liberal side!
It's just rationalization for Newtonian ethics.
https://slatestarcodex.com/2013/05/17/newtonian-ethics/
I have asked two Catholic theologians and they thought Vance was right, Catholic theologians I talk to is not a random sample but I get the impression that the pope isn't seen as an uniquely powrful authority when not speaking ex cathedra.
The pope is still seen as a fairly powerful voice by Catholics even when not speaking ex cathedra.
And both of them voted for him, right?
On the other hand all of the Catholic theologians I've recently talked with are rather unimpressed by Vance theology.
reading about it, i think it may be closer to paul in 1 cor 7 about "it is good not to marry." he is using a negative sense.
the ideal for paul is not to marry and not to have sex as to serve the lord, but to avoid sexual immorality you should marry and in marriage only abstain from sex for brief periods.
the ideal is high but reality is not everyone can without falling into sin. paul would say the highest aim is to be "married to God." Being married is still a love, but not the highest love.
i remember reading that its actually very spiritually dangerous to be a monk for that reason. you have to be able to love responsibility or you become a nut and are worse off than if you aimed lower.
I think the natural place to quote Paul here is instead the times he urged his audience to donate to far off other churches in his epistles
i don't think paul said that as a hierarchy of values though.
2 cor 8 is what i think you mean, and its more "you excel already in other aspects of the christian life, do not forget giving."
he also is pragmatic: their plenty will supply their need, and in turn when you need, they will supply you. the aim isn't to make you poor while others are relieved but to try to equal it out.
i mean i don't think it was held like the concept above was
Re: 10 - I wouldn't be surprised if Musk had multiple people run his account (after all he allegedly hired people to play videogames for him), which would partly explain why he's been able to post nearly incessantly over the past few weeks. The level of sleep deprivation suggested by the tweets/time graph would be difficult to sustain for more than a few weeks, right?
Re: 25 - YouTuber DIYPerks made an "artifical sun" in his home in a way that doesn't seem like it would need to cost $1000/object if it were made at scale and commercialized, and it has the benefit of looking like an actual window. The Brighter lamp is just that, a floor lamp, I'm not sure how convincing the effect would be - like sure, it'll be bright, but it'll still come from a distinct spot on the ceiling and get diffused from above, right? Hardly what sunshine indoors looks like in my experience
> The level of sleep deprivation suggested by the tweets/time graph would be difficult to sustain for more than a few weeks, right?
He is taking drugs...
There is also that one weird gene expression that allows people to operate healthily on significantly less sleep than the norm. I wouldn't be surprised if its fairly common among uberachievers. Modafinil and Adderall only delay the paying of the debt.
I don't like the graph because it squeezes 365 days into less than 46 pixels horizontally. That's less than 1 pixel per week. If he stays up late on some nights and gets up early on others during a week, what does that look like?
(Tweeting just before bed and just after waking up is a problem but one that lots of people have.)
I can't get a handle if his habit is supposed to reach a bad level yet. I thought it's only to show a pattern that it's getting more often
You can also schedule tweets, pretty sure a lot of influencers do that to keep engagement up
I'm still curious how he gets anything at all done with that much X engagement. Not just his own posts, but how much he likely reads from other people! I can't imagine getting anything at all done, let alone what appears to be a busy schedule even for high-achiever's standards.
Hiring people to send his tweets makes a lot more sense to me.
Re: 3 - I heard from someone who worked at the FCC at the time that the reason Net Neutrality ended when it did (and didn't have major consequences) was a result of the shifting corporate balance of power. Big Tech companies had fought and lobbied hard against repeal (partly, I suspect, astroturfing social media) because they saw the ISPs as peers, companies of reasonably similar size who were a real risk if they could take a bite out of tech's pie. By the time Net Neutrality was actually repealed, Big Tech had totally eclipsed ISPs in financial and political clout, and thus knew that ISPs wouldn't even attempt to fight them in the way Net Neutrality advocates feared. Big Tech was happy to shrug their shoulders and let Net Neutrality go simply because they no longer needed it, hence why nothing happened 2017-21.
on the other hand, the internet just gets more and more centralized
I think the damage done is to people who don't want to use big tech
Not really? You still can pirate pretty much every type of content with minimal hassle, something that governments, legacy media, big tech, small tech - everybody opposes. To the extent that people who don't want to use big tech have no good alternatives, it's because their interests are too niche, basically by definition.
#20
>It seems like possibly Trump and Rubio have
The sentence just cuts off there.
Re #8: I think this links to the wrong debate post. In the provided link, Trump doesn't say anything in heroic hexameter.
Just checked - the correct post would be "Hardball Questions For The Next Debate".
Thanks, fixed.
"it’s probably worth it for middle-class people to spend $100 spamming a hundred companies with their applications, and any price high enough to discourage this would make it hard for poor people to apply at all."
But that's fine, though. If the cost of reviewing a crappy application that is an obvious poor fit is less than $1, then you'd be happy to get all the crappy applications people want to send out. Well-qualified poor people aren't hurt by this. In fact, it guarantees companies have the resources to give them a fair shot in reviewing their resume.
In the original, the money is donated to charity. Also, I don't think companies have a lot of slack in trading money for application-review time, since applications usually need to be reviewed by managers who are familiar with the company and its needs.
Currently most applications are reviewed by *software* - isn't this the original problem you described, since the reviewing software favours conventional, less-than-perfectly-honest applicants with the right keywords rather than necessarily good applicants? - which presumably produces some sort of list for managers to pick from.
If the cost of having a human or AI do this first-pass review, rather than a keyword filter program, were less than £1 per application I do think that charging £1 to apply would plausibly address the problem.
I mean, if you have an actually competent AI, none of this would even be necessary in the first place. The cost of having them go through applications is trivial.
Aren't all applications also written by AIs by now? The arms race is going to be legendary! This is exactly the kind of absurd future to be excited about.
Rather than 1$ you could just have them do a 15 minute questionnaire or test.
A lot of places already do this.
Many school districts demand letters of reference, sometimes from a person's current(!) administrator, before even considering their application, electronically or otherwise.
I review hundreds of applications a year. Doing a first-glance approximation cost-effectively at $1/application would be fairly easy. You don't need to be particularly thorough, just separate the "obvious no" candidates from the pool and then pass them along.
You wouldn't want front line managers doing that, but you could easily afford to have additional recruiting staff sift through applications in that manner. It would give you approximately one minute per application, if you're paying an hourly rate of $30 and after loading benefits and additional costs of hiring the reviewer. You could get closer to two minutes per application if you're only paying $12-15/hour.
As it stands, that's a big expense without the $1 per application. Companies getting too many, especially obviously bad, applications need a cost-effective way to sort these applications otherwise, which I guess is where software comes in. My current organization gets around it by having a moderately difficult application process. If we were just going by applications on Indeed we would be overwhelmed sorting crappy drive-by applications from people who don't know or care what we even do.
It's not about paying for the cost of reviewing the application. It's a speed bump to make it harder for literal spammers to spam applications. Much like a captcha. Even my small company has to deal with this. A significant portion of our applications for engineering jobs are from services that file fake applications on behalf of foreign clients. Sometimes the people that comes to the interviews are clearly not the person who "applied". Even a nominal $1 charge would cut down on this only because they would have to provide a payment method. You could probably get the same effect without even charging and just requiring a credit card number.
Re 15. Although I don't believe omega-3 supplementation has any benefit in psychosis, I also don't think this new trial should shift your opinion much, given the total sample size was n=135 and the total number of transitions to psychosis was n=8.
The twin cities of El Alto/La Paz in Bolivia have an elaborate cable car system which is widely used for public transportation (https://en.wikipedia.org/wiki/Mi_Telef%C3%A9rico) -- using it when I visited in summer 2019 was kind of surreal, because you could look down and see the enormous class divides in the city (rich people live down in the valleys where there's literally more oxygen; the flat plain of El Alto is above 4 km in altitude!) while the walls of the cable cars had propaganda for the left-wing government in power at the time, which seemed grossly discordant to me. (But the free speech situation was fine -- our friends there felt totally free to shit-talk the government in those cable cars. This was a few months before the military coup which happened that winter).
Anyway, cable cars for public transportation, incredibly cool idea which has been implemented in some places!
Right, I visited pre- and post-cable-car, and there was a big difference - it's a beautiful system that works. And people felt free to shit-talk the government, which had overstayed its welcome in constitutionally dubious ways but seemed relatively benign compared to previous Bolivian governments (which had a knack for massacring their population, as the same friends and colleagues shit-talking the government also pointed out) and, well, um, let's not talk about what's going on elsewhere.
Agree, I first rode a gondola as part of a public transit system in Barcelona. Seamless transfer from the subway, and it was an unforgettable experience that made me an early believer in Swyft Cities well over a decade later.
Re: 22 - This seems like a classic toxoplasma of rage situation, although with substantially less distinct forms than those in the original post. It's a positive feedback loop, where each side defects harder and harder.
Re: 27 - Maybe some trans-related issues fall here? Although I'm not really sure (the main supporters of the statement "trans women are women" tend to be women, so this goes against that view).
There isn't a massive difference between men and women in terms of whether they see gender transition as morally acceptable. Men are less likely to see it as acceptable, but that's likely confounded by men just being more right-wing. Of course, the biggest predictor of transgender acceptance is just political alignment, to no one's surprise.
https://news.gallup.com/poll/645704/slim-majority-adults-say-changing-gender-morally-wrong.aspx
Re 22: emphatically yes. I think this loop has been running for at least 15 or 20 years in the U.S., probably quite a bit longer. I think changes in media technology and consumption patterns (especially but not exclusively social media) have supercharged it. I don't have much hope that the U.S. will be able to break out of it at this point, because all the obvious tools to do so have already been tightly captured by the loop.
I suspect the fight over “gender ideology in schools” is lower-male vs. upper-female. Education remains female-dominated.
But it might just be sampling bias. I’m exposed a lot more rants from blue-collar men. I had no luck finding polling on CW situations like Florida HB 1557 or the Loudon County policy.
I'm thinking more of the radical feminist version rather than the socially conservative one.
For 1. The mongols used to masacre the populations of cities that didn't submit, and each soldier was given a quota of civilians to execute with their sabre. I expect most cities have larger populations than a mongol army had soldiers. An unusually bloodthirsty mogol that had been at a few sieges must have tallied up a lot of kills by hand.
>The Brighter lamp is 50,000 lumen ... My only concern is that the light costs about $1,000;
This is confusing to me... I already have a 21000 lumen bulb in my floor lamp, that I got on Amazon for $40.
Is there some reason that one 50k bulb would produce 'more light' in some relevant metric than just running 3 of those 21k bulbs?
Or is the price difference really just about the glare/temperature/etc concerns you mention? That seems like an extreme difference in price for just those concerns.
I also have a much cheaper bright bulb. I think it's a combination of glare/temperature/etc, form factor (eg a whole lamp is more expensive than a bulb), and probably pitching to a more premium market of well-off people who don't want to buy their own weird bulbs.
Do these lights come on slowly? How does the eye adjust? I can feel over powered by the relatively bright ceiling lights coming on when I’m relaxing in darkness or near darkness and somebody comes in and hits the switch, but those same lights are not obviously doing anything in bright sunshine. And we can’t stare at the sun, so how comfortable is staring at these lamps if possible.
In my experience, they come on fast, you go 'Aaaargh!' and look away for a few seconds, then you adjust and are fine.
But definitely you avoid staring at them, mine is in the corner mostly behind where I sit, and has a lamp shade diffusing the light and directing it towards the white ceiling.
So yeah it's possible these ones will be noticeably more functional in some ways. Still hard to justify the price difference, but maybe they start as a luxury and the price drops auick as you scale up manufacturing.
Is the point here brightness on dark days, because if I liked brightness at night I’d visit an airport. What I like at night is candles, or that equivalent of lighting. Grey days are depressing though.
The spectrum of the light (related to color temperature but much finer) is important. That costs more. A light that bright in an enclosed space sounds like something out of a black ops site.
Hey, creator of Brighter here. I find my eyes adjust quite quickly (2-5 seconds?), especially if some other lights are on. If you're wondering about dimmability, they're dimmable to about 5% (2500 lumens). You can control this through your smart home app (along with the color temperature.
Specs on amazon are basically the wild west. No one tests anything and almost everything is a lie. Flashlights are notorious for this, with random 6 letter brands competing to see who can come up with the highest number that someone will believe. So unless it's a quasi-reputable brand I would start off being skeptical of the claimed lumen level.
If it _is_ a reputable brand, then all those other factors like temperature etc. probably come into play.
This is the conundrum I've been facing. I would very much like to buy some very bright bulbs, but Amazon is simply no longer to be trusted for electronic items like that. Even if I were to pay extra for a reputable brand name, there's a decent chance I'd receive a Chinese knockoff that might, but probably won't, be as good as the genuine article.
So I stop by hardware stores every so often and take a look a their light bulb aisle. I expect that the supply chains for physical stores are a bit less infiltrated by IP-law-infrigining knockoffs, although perhaps my confidence is misplaced.
>although perhaps my confidence is misplaced.
It is. https://www.justice.gov/usao-ndca/pr/counterfeiters-sentenced-convictions-nationwide-conspiracy-distribute-fake-5-hour
Same. I got a 30,000 lumen work light for ~$100. That said, the glare is pretty bad. I imagine someone handier / with more time on their hands than me could diffuse the light quite easily though.
The one other time I encountered this "lol just use more light, and it works" was on the old SSC so I'm surprised no one referenced it yet.
> Eliezer’s wife Brienne had Seasonal Affective Disorder. The consensus treatment for SAD is “light boxes”, very bright lamps that mimic sunshine and make winter feel more like summer. Brienne tried some of these and they didn’t work; her seasonal depression got so bad that she had to move to the Southern Hemisphere three months of every year just to stay functional. No doctor had any good ideas about what to do at this point. Eliezer did some digging, found that existing light boxes were still way less bright than the sun, and jury-rigged a much brighter version. This brighter light box cured Brienne’s depression when the conventional treatment had failed. Since Eliezer, a random layperson, was able to come up with a better SAD cure after a few minutes of thinking than the establishment was recommending to him, this seems kind of like the relevant research community leaving a $20 bill on the ground in Grand Central.
https://slatestarcodex.com/2017/11/30/book-review-inadequate-equilibria/
#49
> Total forest coverage is still declining
That may be true for the Amazon, in particular, but globally, the pattern is less clear. Wikipedia contradicts itself (https://en.wikipedia.org/wiki/List_of_countries_by_forest_area) starting by stating that forest area increased between 2010 and 2020, but then presenting a table with slightly different figures, showing a decrease.
According to either set of figures, forestry seems relatively stable, with forest coverage increasing over the period by 3.5% according to the first figures or decreasing by 1.2% according to the second figures.
Globally perhaps, but not all forest is the same. Rainforest like the amazon is in swift decline pretty much everywhere. Small comfort if Boreal forest is increasing.
Agreed.
Is there a reason why we need rainforests specifically?
The most commonly cited practical reason would be medicines usually. You never know what Gila Monster is out there waiting to give you the next Ozempic.
For me I'll just say killing the Amazon rainforest feels like a crime against the planet, and against wonder. Like blowing up Everest or killing the buffalo. I believe in the duty to preserve.
I have heard the claim that biodensity is greater in rain forests than elsewhere.
I think that's pretty robust. There are just vastly more species of tree and insect in tropical forests than in boreal ones.
I've heard it's more complicated than that, in an interesting way. Venezuela, for example, is known for its tepui - their version of a mesa. These formations are high enough that their tops are effectively isolated from the surrounding area, and will evolve their own ecologies.
It gets even better. They frequently have natural sinkholes due to the properties of the rock forming them. Those sinkholes can be quite large, and sport micro-ecologies of their own, all isolated from each other.
I get the impression that if something were to wipe out the entire Amazon but somehow leave these tepui untouched, the total number of extant plant and animal species might only drop by a few percent. (I have no known reliable way of precisely measuring this, however - even biologists don't fully know what's in all those sinkholes.)
>The toxic bite of a Gila monster can kill a human, but a specific ingredient in the cocktail of the lizard's venom is the reason we have glucagon-like peptide (GLP-1) agonists like Ozempic and Wegovy.
Huh, TIL.
Evidence over the decades suggests that the Amazon rainforest was itself a human construction for around 11,000 years. (Look up "terra preta".) A rough sketch of the narrative now is that there were as many as 5-8 million people living in what is now the rainforest in pre-Columbian times, perhaps fertilizing and nurturing thousands of square miles of gardens, which ran wild after their numbers declined.
If this is true, then destroying the rainforest might be a crime against pre-Columbian humans at worst.
Humans have lived in and altered the Amazon for about 11,000 years, but the forest itself is over 50 million years old, and at no point did they alter the entire thing. I don't think you could make any claim that the entire Amazon was a human construction at any point, or anything close to it.
How do you know?
Again, look up "terra preta" and related articles. I see no definitive evidence that the rainforest was 50 million years old, and strong evidence that it was densely populated for over 10 thousand years. Presumably, a significant fraction of the land is covered with artificial soil; although no one's been able to perform a comprehensive survey, regions previously believed to be virgin turn out to be growing out of terra preta.
Density and diversity. An acre of rainforest contains a lot more mass of plant life, and contains a much wider variety of both plant and animal species, than an acre of monoculture farmed trees that are be harvested and replanted every fifteen years. But they both count as "an acre of forests".
1) Higher rate of CO2 processing Vs boreal forest due to faster growth (although boreal can have a higher *stock* of stored CO2 per area unit, so it's complicated)
2) More importantly in my view but also more subjectively is the biodiversity in rainforests, which one can make utilisation arguments for (e.g. medicines) but which I think is intrinsically valuable. A universe with an Amazon rainforest is one I value more than a universe with dozens of barren planets or a world tiled solely in kudzu.
A square mile of rainforest holds an order of magnitude more animal and plant species than temperate ones, sometimes pushing towards two orders of magnitude. So, if we were trying to protect only biodiversity as such, we would focus only on rainforests and abandon the others almost entirely.
This is correct. Even if you limit yourself to just the Amazon rainforest, there's so many species that recording and classifying them all is hard. Most of the stuff you hear about "species going extinct before science could record their existence" will be about rainforest species.
A while back I believe the numbers floating around were 25% of Earth's terrestrial species living in rainforests, even though rainforests only cover ~5% of the planet. Now various websites are saying 50% of species, but I'm not sure where they're getting that number from.
The big thing is that we want a diversified portfolio of ecosystems. Preserving one sort of ecosystem has some comparability with preserving others, but it’s better to have some preservation of many ecosystems than to have a lot of preservation of just one ecosystem.
>But where is the symmetrical working-class women vs. upper-middle-class men gender war?
Given that you just used the example of upper-middle-class men not wanting to cheat with the Nanny and considering the Nanny's feelings on the matter, maybe they simply care about women as people much more than the working-class men, and don't feel like going to war against them?
What I found odd about that statistic is I would have assumed only the upper middle classes could cheat with the nanny, given that affording a life in nanny is one of the attributes that defines you as upper middle class.
Well right, but the example (as I understood it) was working-class men telling upper-class women that their husband was going to cheat with the nanny, possibly as a way of degrading the upper-class women for being ugly or career-oriented or not submissive or w/e.
So, working class men projecting their fantasies of what they would do if they were upper class.
The norm was what the working class people believed was true. They don't need to necessarily practice it or experience it.
It could also be sour grapes. "Good thing we can't afford a nanny, because it would destroy our marriage."
It’s the same class. of course the working class guy is going to sleep with the nanny he met down at the pub, if she’s willing and he’s a cheater.
The upper class guy would be sleeping with the help, and that’s considered sordid.
There is also the historic example of women working as servants where they became pregnant by the master of the house or one of his sons, and were speedily kicked out. So working class people may have more memory of what the upper classes get away with when it comes to the lower classes.
https://www.roselerner.com/blog/2020/01/17/being-a-servant-in-the-regency-was-a-truly-terrible-job/
Being pregnant by or having a child with the master would give the servant soft power to wield in the household. This both dilutes the master's power and creates leverage for instability, especially between the servant and the master's wife. The best remedy is to use hard power to eliminate that threat to stability.
Nothing stops the employer from giving the servant a modest sum to live on or get set up elsewhere, but there's also nothing compelling that and it opens up the possibility that it could be used against them later. The master could always claim the servant was let go because she got knocked up and they don't tolerate that kind of immorality. Who are people going to believe, the master, or the servant who did actually have a child out of wedlock claiming it belongs to the master? Giving her money implies she's actually in the right. If you gave every servant who got pregnant money to live on, servants would go pregnant on purpose.
No, it would not. If you're the master's favourite, it will get you better treatment for a while. But a bastard child is not going to be tolerated. The lady of the house will make sure you disappear. The social mores are that pregnancy outside of wedlock is disgraceful, and even more disgraceful if you run the kind of household where the servants are sluts and slatterns.
Some places and times, if there's enough land to carve off the estate, the master may arrange for the pregnant mistress to be married off so some other man can raise the child in exchange for a farm as dowry. That happened round here in the early decades of the last century.
But pregnant maids who may have been coerced or even raped by the men of the household are not wielding any kind of power. Being dismissed without a reference meant you couldn't get a job elsewhere, and a lot of women fell into poverty and did end up on the streets.
On the other hand, upper middle-class couples are the ones vastly more likely to have nannies, and working class/lower middle-class women are the ones more likely to be nannies/working in some childcare role.
So I think upper class women saying "of course Horace would never with the nanny" may be some wishful thinking, and upper class men saying "of course I would never with the nanny, darling" may be CYA.
https://www.reddit.com/r/Divorce/comments/1ejec8a/why_is_affair_with_the_nanny_a_common_thing/
https://www.yourtango.com/2015279705/real-and-unfortunate-reason-dads-go-after-nanny
https://www.girlsaskguys.com/relationships/q5175708-why-do-women-hire-nannies-when-32-of-husbands-cheat-with-them
There arent many working class women that think of themselves as such anymore. Over-education being what it is.
Over here in the U.K. 56% of people see themselves as working class. Which is much higher than the past.
I'm confused by this one, because it isn't remotely difficult to find examples of upper-class (or, at the minimum, fabulously wealthy) men who cheated with the nanny: https://evoke.ie/2025/02/22/entertainment/stars-accused-affair-nanny
Tangential, but class and wealth are not quite isomorphic. The owner of a plumbing company could well be making more money than a consultant at McKinsey, but would probably still be culturally lower class. Movie stars, plausibly also.
(I expect there are also McKinsey consultants who have slept with the nanny, so I'm not engaging with that part of the argument).
Yes, I agree that this is a distinction that matters. Ben Affleck is probably culturally working-class, so it's perhaps less surprising that Ben Affleck slept with the nanny than some guy who went to Harvard on a legacy admission.
On the other hand, Ethan Hawke's father was an actuary and Robin Williams's father was a senior executive in Ford; both apparently enjoyed quite privileged upbringings. It seems reasonable to describe both men as culturally middle-to-upper-class.
> Ben Affleck is probably culturally working-class, so it's perhaps less surprising that Ben Affleck slept with the nanny than some guy who went to Harvard on a legacy admission.
You have a very warped view of classes, culturally or otherwise.
It’s not on the least surprising when Harvard graduates are cads, it’s a solid working class guy like Ben Affleck you’d expect to be decent.
( actually on googling him I see his mother was a Harvard educated teacher, so maybe that explains the blackguardism).
>You have a very warped view of classes, culturally or otherwise.
>It’s not on the least surprising when Harvard graduates are cads, it’s a solid working class guy like Ben Affleck you’d expect to be decent.
My comment was quite clearly made in the context of a discussion about an article in which a woman argued that working-class men think that cheating on your wife with the nanny is inevitable. If my view of classes is "warped", then so are the author of said article.
PEPFAR is great and should continue, but I don't think making claims like this is helpful: "PEPFAR Impact Counter tries to estimate the number of people affected, and says that 13,854 adults and 1,474 infants have already died from this policy."
Unless I'm missing something HIV typically takes years to show any symptoms, much less cause of death.
Good point. I'm not an expert in this, but the methodology section of the counter https://pepfar.impactcounter.com/methodology says things like:
- "The recovery of the immune system among individuals living with HIV through antiretroviral therapy can take several years. However, this recovery is quickly compromised when the virus begins to replicate in the absence of treatment"
- "Of the infants born with HIV who do not receive treatment, 20% are expected to die within the first three months of life "
But I don't have a good sense of how much these are deaths that have happened, deaths that should be expected in the future based on things that have happened now, vs. the amount of deaths you would get if you suspended PEPFAR by X times the amount of time it's already suspended, divided by X. My guess is some combination of all of these.
Do you actually need to see anybody die in order to produce a reasonable estimate for the number of people who have?
Presumably we know roughly what the mortality rate is in populations supported by PEPFAR, and in populations unsupported by PEPFAR; if the difference in those two mortality rates (multiplied by the population size and by the time since PEPFAR was cancelled) works out at 15000-odd people, that seems like the main thing we need to know?
(In fact, unless I'm missing something too, it seems to me to very much be preferable to work out an estimate in this way rather than waiting for those 15000 people to actually die before starting to try to count them..!)
I think we expect there to be a several month lag between treatment starting/stopping and mortality rate changing.
Yes, I understand that - but if we have very good data about what the mortality rate is going to change *to* (for example because we can see the mortality rates in non-PEPFAR-but-otherwise-demographically-identical populations) then surely we don't need to wait for it to actually change before we produce our estimate for the number of deaths?
(I'm fact, if we're trying to figure out whether this policy change is worth the number of deaths it might cause, I think it should be basically essential to produce the estimate before the mortality rate changes! If we're trying to calculate out whether or not we can afford to help some people it hardly seems useful to only start doing the maths after they're dead!)
> 13,854 adults and 1,474 infants have 𝗮𝗹𝗿𝗲𝗮𝗱𝘆 died from this policy.
Okay, sure, it would technically be more accurate to say "the deaths have been irreversibly set in motion but have not-yet actually occurred". I think "have already died" is a fairly common colloquial way to express that in a morbid-but-pithy sentence, but I can see this being inappropriate in an academic work (or else an indication that maybe the author had the same misunderstanding I had*)
* Your emphasis really helps me there - thanks! - I thought the argument was about how early you can reasonably make the estimate rather than, given the estimate, how early you can say "have died")
I think the irreversibility is also somewhat dubious? At least, much more dubious than if they were in coffins, as is pretty clearly implied.
It's an awfully specific number as well. Saying "approximately 14,000 adults will die/have died from this policy" doesn't imply as strongly that you know the names and circumstances of each individual who died and can directly confirm that the death was from lack of PEPFAR funds.
I suspect the people who produced that number are very well aware of the difference, and are willing to make it sound more concrete in order to gain that extra leverage.
I don't think they're immoral to do so, but I do think we should discount what they say much more because they are willing to do that.
In some of these cases, I think it’s also likely that a three week gap leads to zero deaths, even if a twelve week gap leads to a large number of deaths.
>Related: this is all fun to think about, but very early polling for the 2028 Democratic primary suggests that by far the #1 candidate is . . . Kamala Harris at 37%, beating Mayor Pete, Gavin, and AOC with 11%, 9%, and 7% respectively. I know you’re not supposed to take early polls like this seriously in terms of who will actually win, but can you take them seriously as a guide to whether people have learned any lessons / no longer love losing?
Take it as a lesson on why first past the post voting is absurd, and should be abolished!
Harris is winning that poll is indeed on name recognition alone, which is how a lot of people win first past the post elections with lots of candidates (plausibly how Trump gained momentum in his fist primary, for example).
Try again using Approval voting, you may get very different answers.
Doesn't high name recognition also lend an advantage in approval voting? While some people do just vote at random, I think in general a voter would be less likely to approve a candidate they don't know well.
It's still an advantage, but much less of one.
First of all because if you approve one person because you know their name and 5 other people because they seem alright after you listened to them talk for 5 ins (o w/e), they all get equal support form that. Vs voting for the one person with the big name and everyone else gets zero support.
But, second and perhaps more importantly, because it overcomes the issue of strategic voting through Schelling points. A big reason that many people support high name-recognition candidates is because they expect less engaged voters to know who that person is and rally behind them, so it feels safer to vote for the well-known person to avoid intra-party infighting and lost momentum. With Approval you can still approve that person in case that is what ends up happening, but also approve everyone else you like and give them a much better chance of competing.
Hillary Clinton had name recognition that's about as high as possible to achieve, but she still lost. Her negatives were huge.
I am not that immersed in American politics so I'm actually curious why they don't just fucking run Hillary again. Harris now has the exact same record as her against trump!
If Hillary were 60 years old instead of 77, they probably would.
“They” don’t run anyone. The person themself has to enter the race. Everyone who says “why don’t they just…” doesn’t understand how American political parties work.
Early polling is always like this. Same as Vance being number 1 on the other side.
Amen. My single issue.
Unfortunately, my local legislators have suggested it’s unreasonable and possibly unAmerican.
FPTP has the advantage that it is simple to administer and simple to understand. We already saw how much of a shit show things can be with people not understanding how elections operate. It would be even worse if you use some fancy voting system that doesn't translate to results in an obvious way.
Approval is about as simple and easy to administer.
'Vote for everyone you like, most votes wins' is actually fewer words than 'vote for the person you like most, most votes wins. And you can use existing ballot and voting machines with no overhaul, just let people fill in multiple bubbles.
Re: 22, I don't share your views on DEI but I still struggle to steelman them into a view of their impact on federal workers that produces anything close to the harm of the doge cuts. Just look at the Ted Cruz survey you posted about -- some studies you don't like but they were a fraction of a fraction.
Not sure what you mean - I would guess a significant number of federal hirings and promotions (~25%) were impacted by DEI concerns, there was an additional chilling effect where white men wouldn't apply, and even before that the federal hiring process was retooled to give better DEI results. See eg https://www.tracingwoodgrains.com/p/the-faas-hiring-scandal-a-quick-overview. See also section 3 of https://www.astralcodexten.com/p/highlights-from-the-comments-on-the-cf9 , especially Martin Blank's (though notice that some of the comments above contradict)
It seems to mee that the FAA example is at an extreme and was recognized and addressed by the system eventually. Also there is lots of evidence of systemic discrimination against non-white applicants in many fields (not specifically the government as far as I know) so DEI policies are doing some good even if in some cases they are also causing harm. Even if 25 percent of hires are "impacted" by DEI considerations, Musk is talking about straight up firing far more than 25 percent of federal workers, though he hasn't quite gotten there yet.
I think it's implausible that removing discrimination helps more than affirmative action hurts. The most easily measurable thing here is college admissions, and we know that admitted minorities got much worse SAT scores and other measurables. This is also my impression of the Get More Women In Tech movement. It's possible government is an exception, but I haven't heard any good arguments why that would be.
I think lowering the quality of workers probably hurts more than lowering the quantity - if one general makes bad decisions, that's plausibly worse than having 25% fewer soldiers. But I agree that if Musk fires "far" more than 25% of federal workers then it will become implausible that DEI had a larger effect.
I think both Musk and DEI will have their largest effect in very smart people with other options realizing that it would be masochistic to go into public service, and sticking with Wall Street or something instead. I think government needs some of those people to function well and could barely do a better job kicking them out if they tried.
I guess I have a hard time seeing how you get from *any* of these things to the chilling impact of putting forward indiscriminate cuts as a possibility. Even if they stop tomorrow and don't fire anyone else, they've already decimated the federal hiring pool for many years, until some Constitutional change can credibly promise that this sort of thing won't strike again.
That's a good thing.
Why is it a good thing that we have lost one tool that we use to hire good employees? Is it better to pay them higher wages instead?
I don't follow. The boss of Google can lay off any Google employee at any time, but Google can still attract talent. My impression is that talented people are willing to work at places where they might be laid off as long as the chance is low. I've heard claims that DOGE wants to fire ~10%. Suppose that every Republican president does something like this and Republicans win half the time - a 10% chance of getting laid off per eight years doesn't seem that different from private industry.
There is the obvious difference that Google pays a lot more (in dollars) than the government does. In the past the government was arguably paying partly in security, but the government total compensation package has just received a drastic cut. You would expect that to hurt recruitment, unless they up their pay in actual dollars to compensate.
The head of google doesn't change every 4-8 years into someone who hates all the employees the previous head hired.
Along with the point about government jobs compensation package including less money but more security, the general assumption at private companies is that during layoffs management attempts to dismiss only their poorly-performing employees. So you can avoid being laid off by working hard and being good at your job. Obviously in the real world no one has solved the problem of accurately measuring every employee's value/quality, but at least they try.
The DOGE cuts have conspicuously not even attempted to fire bad employees and retain good ones. Federal government departments do have HR and managers and annual reviews and such just like private companies, and one would think a bunch of programmers whose skills are sufficient to fiddle with the US Treasury's money disbursing database could also analyze the HR databases, but here we are...
Google has in fact become, from my social circle's anecdata, a significantly less desirable place to work since they started doing mass layoffs in 2023 for the first time. People were particularly mad about the lack of transparency around the layoff decision making process and the lack of obvious correlation with performance. It wasn't as bad as DOGE in those respects, but it reflected an arrogance towards the rank and file that seems unfortunately to have gotten more common even among non-insane tech executives over the last 5-7 years. The rank and file are not blameless in this-- the cartoonishly dumb woke techie activism of the late 2010s rivaled that in universities-- but the phenomenon is nonetheless very bad and likely has eroded morale and productivity in much of the industry.
This misunderstands the (to date) compensation structure of Federal employees, including some of the uniformed service members effected by DOGE decisions and DEI purges.
Generally, Federal compensation is heavy on non-salary compensation. Historically, this has included defined benefits pensions which vest after long periods of service, but that's mostly been phased out these days in favor of other things. Recently, this has also included flexible telework options for most employees. You also get a measure of job security not enjoyed by most private sector folks and often had opportunities to directly effect unique missions that weren't accessible in the private sector. In exchange, you get paid less than you would in the private sector. In some areas, this difference in pay is so large that the Federal government has struggled mightily to recruit talented individuals.
There was an understanding (varying by degree of how politically-facing your office was) that you would be insulated from the day-to-day shifting of the political winds. For example, the Commissioner of BLS might have different priorities depending on who appointed them, but those changes were mostly around the edges, and the core mission of collecting and publishing credible data on the market wouldn't change.
And for the most capable employees, a crucial part of the compensation package was the mission. Private companies can also have appealing missions, but there's a lot of pro-social work that can only be done at the Federal level, and for many Federal employees (especially the best, who could easily find other, more remunerative jobs in the private sector), that was the primary appeal. Or at least a key component that compensated for some of the other, less enticing aspects of working for the Federal Government.
These past two months have struck directly at those non-pecuniary aspects of the compensation package.. Apparently, you can be fired capriciously by some rando Presidential Advisor. And you'd better get damned political, because the current party thinks every Federal job is a political appointment, and damn your mission.
The approach DOGE is taking is to eliminate the professional bureaucracy that served the missions of the agencies they were a part of. It takes as a starting point that bureaucrats should serve the President first and their agency (and their agency's missions) second. This is part and parcel with the a "unitary executive" legal theory the current administration is aggressively pushing. I have no opinion on its merit w.r.t. the Constitution, but its effect on the professional bureaucracy and the independent agencies they support will be devastating.
Surprised by this take. I'd much rather have a fully staffed IRS than a 80% staffed IRS with slightly more competent people on average.
Also, talented people who choose to take government jobs instead of industry when both are available do so because of stability. These people are very risk averse and defaulting on them once (or just laying off anyone) is really bad for attracting risk-averse and talented people in the future.
I wonder if that's why government norms seem to be way too risk-averse to be compatible with sanity. (Classic examples: FDA approval, Covid challenge trials)
These are decent examples, but for a lot of function of government, risk-aversion is a good thing. The worst outcome of taking risks as a private company is (usually) that the company fails. The worst outcomes of taking risks as a government responsible for hundreds of millions of people make entire chapters in history books.
>It seems to mee that the FAA example is at an extreme and was recognized and addressed by the system eventually.
When was it addressed by the system?
I’m having a bit of trouble following the details, but it looks like it was started in 2014 and stopped in 2016, though there was some long-lasting damage.
https://www.tracingwoodgrains.com/p/the-full-story-of-the-faas-hiring
There was a law passed by Congress. Trace covers all of this in his articles on the topic.
He doesn't seem to think that the law fixed it, since he goes on to say "The Trump administration missed it, too, for a term, and it’s likely most officials simply didn’t hear about it through the first few years of the Biden administration." and "I badly wanted the Biden admin to rise to the occasion, speak plainly and frankly in response to the scandal, and defuse this time bomb." and "the Biden admin declined to fix it" and "Democrats had a chance to clean it up, and they didn’t", all of which are about events after the law was passed. So if Trace is your source, he's also the one saying that the 2016 law did not fix things.
The 2016 law did not provide recompense to the affected employees, but it did eliminate the hiring practice going forward. So it all depends on what we mean by "addressed by the system." There's an ongoing class action lawsuit seeking that recompense (and we only know the incredible details here because they pulled those details out during discoveries over the last decade.)
> Also there is lots of evidence of systemic discrimination against non-white applicants in many field
Is that actually a major problem when you have plenty of competent white applicants to choose from?
This adds nothing to the conversation. This is the equivalent of just responding "no, you're wrong." Your signal to noise ratio is dire.
No, he's arguing that DEI is good because it removes discrimination, I'm arguing that the effect of discrimination on productivity in negligible because the majority of competent applicants are whites, due to demographics and... other reasons. It's basically the same argument Scott is making above.
He has a whole host of reasons for opposing discrimination and you are just saying “discrimination is fine.”
There is also lots of evidence of systemic discrimination against men, favouring women, and yet we were constantly told the opposite, and institutions of power pervasively engaged in overt discrimination (like women only fellowships) in addition to the covert one. Somehow gender discrimination gets less attention than racial one, but there's a solid case for men to be angry.
Sure, but two wrongs don't make her right.
Taking that number at face value, is 25% of federal hirings and promotions over a four-year period really all that big? If (at a wild guess) the average person who works for the federal government at all in their life puts in 20 years in total, that's only ~5% of hirings that would be affected (I'm less sure how to guestimate promotions).
Meanwhile "impacted by DEI concerns" seems like quite a low bar. How much impact are we talking about here? Intuitively, it seems like this would be the biggest issue in cases where the qualified applicant pool is very small (so the value-over-replacement of the best candidate is likely to be high), which is *not* my impression of what most important federal jobs are like.
Also, pro veteran bias is *much* stronger than any DEI stuff. It's common to see hiring slates that are 100% veteran.
DEI doesn’t seem to be particularly harmful to private industry. McKinsey has put out four reports showing that companies with more diversity on their boards and in top management perform better. The fact that they got similar results in different years, and while more than doubling their sample size (so they weren’t just repeatedly looking at the same companies) suggests that this is not just a statistical fluke.
I found the following article that pushes back against this:
https://knowledge.wharton.upenn.edu/article/will-gender-diversity-boards-really-boost-company-performance/
The article correctly points out that correlation does prove causation. It also points to academic research which shows that “the relationship between board gender diversity and company performance is either non-exist (effectively zero) or very weakly positive.” It cites a finding of a minimal (though statistically significant) correlation between gender diversity in top management and corporate performance of 0.03.
Perhaps DEI program harm corporations in way that these studies can’t pick up on because the studies only look at diversity in top level management and on boards of directors, ignoring lower level employees. It is also possible that a DEI program that is harmless in a corporate setting will have a different effect in a government setting. Still, I’d need a bit more than anecdotal evidence to conclude that DIE programs were doing major harm to government efficiency.
Ah yes, McKinsey, known for their dedication to unbiased knowledge-seeking and absolutely not for shitting out 100 page slide decks supporting literally any position a client is willing to pay for.
I would add https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3849562 to the Wharton link.
This paper, contra McKinsey, doesn’t find a statistically significant benefit to diversity in top management. I’m inclined to trust it more than McKinsey.
The question I was considering was not whether DEI helps financial performance, but whether it hurts it. This paper doesn’t find a statistically significant effect, which leaves open the possibility of a negative effect that is too small to detect without a larger sample size. However, the Wharton piece references a metastudy that did find a very small but statistically significant positive correlation. So I’d say the evidence still points towards no negative effect.
Thank you: I'm not a *fan* of a lot of the DEI stuff, except to the extent it accelerated the trend of people not being openly racist and sexist on the workplace (still a thing ten years ago), but the harms seem wildly overstated outside of academia.
... and people don't see how much DEI in academia is by gender and how little is by race, because most people's experience of academia (if they have any) is centered at the undergraduate level, and thus gives much more weight to undergraduate admissions than to anything else.
https://econjwatch.org/articles/mckinsey-s-diversity-matters-delivers-wins-results-revisited
Abstract
In a series of very influential studies, McKinsey (2015; 2018; 2020; 2023) reports finding statistically significant positive relations between the industry-adjusted earnings before interest and taxes margins of global McKinsey-chosen sets of large public firms and the racial/ethnic diversity of their executives. However, when we revisit McKinsey’s tests using data for firms in the publicly observable S&P 500® as of 12/31/2019, we do not find statistically significant relations between McKinsey’s inverse normalized Herfindahl-Hirschman measures of executive racial/ethnic diversity at mid-2020 and either industry-adjusted earnings before interest and taxes margin or industry-adjusted sales growth, gross margin, return on assets, return on equity, and total shareholder return over the prior five years 2015–2019. Combined with the erroneous reverse-causality nature of McKinsey’s tests, our inability to quasi-replicate their results suggests that despite the imprimatur given to McKinsey’s studies, they should not be relied on to support the view that US publicly traded firms can expect to deliver improved financial performance if they increase the racial/ethnic diversity of their executives.
You have no evidence for that 25% number and even if it was true, do you have evidence that the people hired were of less quality? If you have two candidates of equal quality, and you hire the black one because of DEI, thats a "DEI hire" but it doesn't impact the quality of work.
I suggest reading the TOOW review. "Only when two candidates are exactly equal" was always a fig leaf - the law demanded that companies massively change the ethnic composition of their workforce in a way that they couldn't do while only hiring exactly equal candidates unless you grant implausible assumptions about how often there were two exactly equal candidates of different races applying for the same job. This is also not the experience of hiring managers I've read commentary from.
I read the review when it first came out and skimmed it again just now to refresh my memory. Nothing in the review provides any actual evidence on the impact of supposed DEI hiring on the effectiveness of the federal workforce.
Data that I would like to see:
How many people got jobs because of DEI policies they wouldn't have gotten without DEI policies.
The people hired through DEI policies are less effective than those not hired by DEI policies
Heck, I would take any objective data on the federal workforce being less effective now than in the past!
"law demanded that companies massively change the ethnic composition of their workforce"
I thought we were talking about the federal government here?
Anyway, which law specifically? "read the TOOW review isn't a good answer, there are lots of civil rights and EEOC laws. TOOW presents a long chain of inferences and assumptions about what may have happened. But where is the evidence that it actually has?
If DEI has been such a horrible policy, then it should be trivial to present data about it. Yet I haven't seen any! And I ask for it in these threads all the time!
"(~25%) of federal hirings and promotions were impacted by DEI concerns" can only be justified by an absurdly weak interpretation of "impacted" or an absurdly broad interpretation of "DEI concerns".
The FAA scandal is an excellent example because it is simultaneously (1) egregious, (2) exceptional, and (3) per TW's own assessment, had minimal overall impact on the quality of ATCs who made it out of the pipeline. Revelations like that should cause you to believe that DEI has relatively small effects on hiring and promotion decisions, not large ones.
The bigger thumb on the scale by far in federal hiring is veteran preference
I had the same reaction to this bit, while generally agreeing with Scott's pessimism re: the admin's cost-cutting approach. Specifically:
"I think the main effect will be saving ~1% of the budget at the cost of causing so much chaos and misery for government employees that everybody who can get a job in the private sector leaves and we’re left with an extremely low-quality government workforce. I freely admit that DEI also did this, I just think that two rounds of decimating state capacity and purging high-IQ civil servants is worse than one round."
This assumes that DEI practices left us with "an extremely low-quality government workforce," which seems to me like an empirical claim not really supported in Scott's responses here besides anecdotes on hiring from the comments on his Hanania book review and an analogy to college admissions and SATs. Is there output-based evidence suggesting this is true? May be I'm unduly assuming government work is an exception of some sort.
FWIW, from what I've seen and heard (D.C. lawyer), in the legal field you're already choosing from a slimmer talent pool with a GOP admin because lawyers trend liberal and many don't want to work for it (this may be a Trump era development, not sure). I suspect the Trump 2.0 politicization of the administrative state likely will exacerbate this problem. Plus we've already seen them drive away even very conservative senior staff with tomfoolery (see the Eric Adams deal and subsequent resignations). So firing experienced staff otherwise willing to stick around may be especially harmful for competence in legal positions.
Why would you even need competent legal staff when you're already above the law?
The SC rules that only Trump personally is above the law (in the course of performing his official duties). This doesn't apply to subordinates and you can only pardon so many people before the Presidential pardon will be taken away by Congress. (I wouldn't be surprised if that happens in the next few years, in any case.) It's also a problem that the pardon only plus to violations of Federal law.
Biden could have become the ultimate based president if he had pulled out a gun and shot Trump during their post-election meeting. According to the insane logic of the Supreme court, it would have been legal or at least in a mere legal Gray area. Maybe Trump can use that one weird trick on the president elect if the Dems win in 2028.
> This doesn't apply to subordinates and you can only pardon so many people before the Presidential pardon will be taken away by Congress.
...The congress that's controlled by Republicans that stood behind Trump after Jan 6 instead of throwing him under the bus? That continues to stand behind Trump as he commits an ideological purge of federal agencies and destroys old alliances to ally with a dictatorship? These people are absolutely in it for the long-haul. As for the mid-terms... there are an endless amount of ways to deal with that when the law no longer matters.
You're fantasizing about some kind of revolution. But Trump isn't a revolutionary figure and the Republican party is not a vehicle for revolution. The Republican party is fundamentally conservative. It's not going to allow Trump to just sweep away all laws in order to do a revolutionary overhaul of the country. That's just a fantasy. Even Trump supporters have their limits. Notice how they boo Trump when he talks about the beautiful vaccine that he created. Americans believe in the law, and Republican voters care about the rule of law as much as anyone.
Hell, disorder and lawlessness is a big part of the reason why they have rejected the democrats. A majority think that January 6th was a false flag executed by antifa members. You've been reading a lot of liberals complain about how Trump is a revolutionary Maoist who is going to totally upend everything, and I'm afraid you've gotten your hopes up. Are you even American? It really sounds like you interface with this country indirectly, not directly.
It's easy to stay detached when you have nothing to gain and nothing to lose. Regardless, I doubt I'll be able to convince you, and there isn't much point in doing so in the first place when there's absolutely nothing you can do to stop this. So please... just enjoy the show.
I have seen no evidence that "DEI hiring" has caused any impact on federal workforce effectiveness. Its astounding to me that Scott and so many "rationalists" just take these claims as true without any evidence of it.
Re 37:
Is AIs' refusal to talk about topics like race/IQ not just the same kind of "woke" opinion that's explained by the liberal/conservative divide in training data?
It's hard to get refusal to talk out of training data, because AIs are specifically trained to want to answer all questions, so you have to specifically train them out of it. If they'd just given a typical liberal answer on race/IQ, I would believe it was just training data. I think AI companies chose not to go that route because you could always prompt them differently to give a non-liberal answer, unless they were trained otherwise.
Whatever podcast is so bleak. Just this dude trying to catch out girls who aren't even old enough to drink in a logical inconsistency so he can feel smug about them because he knows he'll never get to fuck them. Although admittedly the Gorlock memes were pretty funny.
I've never heard of this podcast before and can scarcely imagine anything I might possibly care to watch less (also: since when did podcasts become something one *watches*?) - but if 'catching beautiful women in logical inconsistencies during debates' is this geezer's kink, who are we to judge?
>since when did podcasts become something one *watches*?
All of the episodes are uploaded to YouTube as well as streaming platforms, just like Joe Rogan.
Andrew Wilson is...certainly something. My cursory understanding is that he also argues against democracy with randoms brought onto these podcasts. It's the most annoying form of engagement, but it gets clicks, so Whatever, I guess.
Point 5, to me, sounds a lot like AIs having either a virtuous or vicious moral character.
Letby is 100% guilty.
Argument?
There is a long debate about the subject on the subreddit
https://www.reddit.com/r/slatestarcodex/s/I9DncNCbAq
I found the evidence presented by Spiked to be enormously damning:
https://www.spiked-online.com/2025/02/08/the-devils-advocates/
https://www.spiked-online.com/2025/02/11/no-lucy-letby-has-not-been-exonerated/
https://www.spiked-online.com/2024/08/30/the-case-against-lucy-letby/
https://www.spiked-online.com/2024/07/10/lucy-letby-is-guilty-get-over-it/
Thanks for sharing, I will read these, I also think Letby is probably guilty but Private Eye, who have a good record on miscarriages of justice, think otherwise: https://www.private-eye.co.uk/special-reports/lucy-letby
I'm not sure that Private Eye does have that great a record. They throw a lot of mud on the wall, and some of it sticks, but plenty of it is quietly forgotten.
They went all in on supporting Andrew Wakefield, the doctor who linked vaccines to autism.
Banned for a month for this comment; try not to raise the temperature unnecessarily like this.
They do some excellent journalism and book reviews. But their editor, Ian Hislop, is a fanatical Europhiliac, who never misses an opportunity to knock Brexit! That spoils the magazine for me.
The biggest piece of evidence is the fact that she was found guilty in court.
All of the big expert claims and long debunkings are classic "man of one study", although it might be better termed man of one claim.
They go through the available public evidence, find a specific medical or other technical claim, then go expert shopping to find a group of experts who disagree (which as you no doubt know for medical claims, is never very difficult). Then, having found enough doubt to cast on this claim, hey presto, Letby must be innocent.
It might well be that the pro-Letby experts are entirely correct in some cases. But, no reference is ever made to all the other details of the case which were used to convict Letby. There was a lot of traditional police work that went into her conviction that is ignored
The expert leading the panel debunking literally wrote the paper the prosecution experts leaned on. That's hardly shopping around.
This was an expert they got well after the initial case, when they presented other experts, and after other "expert panels" ruling on other minutiae had already been presented and failed to make the desired impact.
It's hardly minutiae when the expert wrote the paper the prosecution used to propose a murder method. And when that same expert says they misused his paper and the baby could not have died in the way the prosecution said.
A lot of studies are misused. It means nothing.
I personally believe that she is guilty but retain 5% probability that the evidence is faulty. My suggestion would be instead of guilty or not introduce another category – probably guilty – which is not enough to jail someone but enough to prevent to working with children.
I have to ask then why was Letby charged. Was she made a scapegoat for the failings of the hospital care? Was it just a coincidence that she started working on wards where the mortality rate went up? Bad luck? Why her and not someone else?
There have been enough killer nurses that I suppose there's a certain amount of hysteria at work. And Letby was just in the wrong place at the wrong time. Assuming she's innocent, which I believe, although I'm no expert on the case.
If Letby is exonerated on a "reasonable doubt" that doesn't prevent the NHS refusing to employ her, her professional body refusing to let her practice or Child Protection measures being taken against her.
Christopher Snowdon has a few articles on Letby that the ACX reader might enjoy (e.g. https://snowdon.substack.com/p/lucy-letby-and-the-texas-sharpshooter ; https://snowdon.substack.com/p/is-lucy-letby-innocent). I haven't gone into the details of the case, but this quote seems to capture the argument that overwhelming circumstantial evidence and lots of very suspicious behaviour points to her guilt:
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
The case might be unstable in a few of the medical areas, but it seems unlikely that this is all just an unfortunate coincidence for Ms. Letby.
Re: gondolas. I spent a season as a ski bum pre-career, with the interesting wrinkle that I specifically worked on what was at the time billed as the only gondola-based municipal mass transit in the US. AMA I guess, and some thoughts:
- Gondolas obviously have a huge advantage in mountainous terrain where other mass transit is a non-starter. I'm not sure the economics will pencil out otherwise given the fairly limited throughput, though some designs can get to some fairly impressive capacities.
- The individualization of compartments is both a blessing and a curse, overall I like the trade. The consistency of service and increased privacy is quite nice off of peak hours. People can and do make messes in the cabins, with vomit being by far the most common (same issue as Uber/Lyft, for people coming back from the bar). Hard drug use was nonexistent, hotboxing was common - though that might have been an artifact of this being Colorado in 2014. It's trivial to take an individual cabin out of service for cleaning with minimal disruption.
- This is a high-contact service position; operators *have* to know the local PD has their back. Standard practice for egregiously disruptive members of the public was to let them on the next cabin, then punch in a stop once they're in the air until PD can meet them at the next station. Striking an operator is like 3 different felonies, DON'T do it.
- You will need to pay operators above local prevailing wage if you want folks to pass a drug test and get that sweet Federal/State funding. They *will* fail if you retest often though, see hotboxing note above.
- All the various issues of "how do we accommodate people with XYZ disability" have been solved, but in most cases it results in a 30-60 second delay for everyone on the line.
Interesting observations. Where was this?
Telluride / Mountain Village, CO, 2013-14 season
Ah, now that sounds familiar Was definitely an interesting experiment, and appreciate your observations & experience about it.
Re #6 l, the main problem is how little signal you could ever extract from a resume. I/O psych research generally favors intelligence tests, structured interviews (although some researchers still apparently think unstructured interviews can be solid), realistic tests of job skills, a small number of personality dimensions, and experience (but only in the 0-5 year range), as good predictors. Educational history is weak, references are weak, experience over 5 years is weak. The only clear signals you can pull from a resume are "does this person work in this field?" and "have they for 5 years?"
Beyond that I'd be really suspicious of any AI system that supposedly does a good job; it's probably just figuring out how to reflect the biases of the people using it.
Re 3 - your take is mostly true. My career background is in engineering at large ISPs. so I have significant understanding here, although there are people in the industry that have better grasp on the regulatory side. A caveat is that I've never worked at a residential provider, but I am an industry expert on backbone topology.
Without getting too deep in the weeds:
1. At the time of the second big push for Net Neutrality (roughly 2014-2017), most ISPs backbones were on speeds and/or protocols that could be considered two grades behind where things are currently and had been relatively static on this front since roughly 2006. The 2014-2017 period coincided with when adoption/utilization of streaming services, which is unequivocally the biggest driver of Internet bandwidth, really started to uptick.
2. The content providers, even those that were early adopters of CDN approaches, realized around this time that they couldn't continue to grow if user experience over the insufficient average home Internet connection of the time continued to be poor. In addition to pressuring the major providers to invest in their networks (often by spending money directly with them for this purpose), the largest content providers also radically rearchitected their own environments in order to reduce the impact that a user's ISP architecture would have on end user performance.
3. Continued investment of federal funds into broadband improvement - while there is plenty to be aggrieved about with the implementation of nearly all of these programs - has sustained continued improvement in the average home Internet user's experience through all subsequent Presidential and, therefore, FCC administrations.
Notice I haven't talked about net neutrality or the regulatory environments at all.
While the previous statements are facts, the following statements are (I believe to be highly informed) opinion:
1. Some of the things Net Neutrality proponents warned against are not viable, because (most) ISPs are capital-constrained and extremely slow to innovate, even (often especially) when the larger parent companies are massive money-printing machines, and would require them to be nimble and agile.
2. Some of the things Net Neutrality proponents warned against are viable and definitely deployed by many ISPs but have little to no adverse impact to end users. Zero rating certain traffic is a good example of this.
3. Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
4. Net Neutrality is actually a red herring - on the Internet, treating traffic equally and treating traffic neutrally are different and both principles violate reasonable traffic management practices. The main reason for this is that Internet backbone routers are essentially armed to make their own decisions, and while the parameters to make those decisions can be tightly controlled, the end result of those changes can be variable, even when done by seasoned veterans in well-controlled circumstances (and often neither of those things are the case). Lack of customer choice due to local monopolies or duopolies is a real problem (which emergent providers and technologies are slowly addressing) but Net Neutrality is a theoretical argument, and customers care about user experience not philosophy.
5. This is slightly less informed opinion, as I have never worked for a residential ISP, but I don't believe that (most) ISPs are agile enough to change their traffic management practices based on regulatory environments that can potentially change every four years federally (let alone state-by-state environments) and therefore not responsive to it at all.
The zero rating thing is the reason why in my own (admittedly much less technically informed opinion), that I said I thought lack of net neutrality is probably actually having small, negative impacts.
Zero rating specifically benefits large incumbents who can make those kinds of deals. It therefore dis-incentivizes new entrants.
Theoretically, we should expect a reduction in competition because the existence of zero rating presents a new barrier to entry/competitive friction.
As I also mentioned, I expect this effect to be relatively small and hard to see, definitely much less severe than what the doomers were saying at the time. "
I actually agree with you about zero rating in theory. In practice, however, I believe it to be a non-issue. My guiding principle is always "is the end user experience adversely impacted?" and as zero rating is a commercial practice and not a technical one, it requires end users to make different utilization decisions specifically on account of some content being zero rated in order for the answer to that question to be yes. Industry data that I've seen consistently shows that this isn't happening. That said I firmly believe that the increase in utilization caps for home Internet and mobile plans, along with topology changes on the content side, have made zero rating more of a non-issue for end users than anything else - in other words, traffic not counting for customer utilization doesn't really matter in that environment. I'd like to see some specific data for users that are on more tightly constrained plans but it's such a small portion of the overall user base at this point that I'm unclear if the data would be meaningful.
The other issue you point out is creation of a barrier of entry on the content side. Again, on paper, I agree, but in practice there are simply much bigger barriers of entry. On paper "start up creates competitive search engine to Google and gets blocked out because end users can use Google for 'free' but not the start up because they can't afford to pay off Verizon" sounds bad, but the billions of dollars the start up would need to create infrastructure that can support creating a product competitive with Google's is orders of magnitude larger than whatever deal would be struck with a carrier. Moreover, so much of this type of new entrant lives in hyperscale cloud infrastructure that this isn't really where zero rating comes into play anyway (in part because the level of granularity required for a carrier to differentiate between traffic sourced from, for example, within the same AWS-owned IP block, lacks a cost benefit), it's for the "competitor to Disney+" or "competitor to Spotify", services that source much more data in aggregate and in wholly different ways than a more static website, and there a whole boatload of bigger reasons why competing with those services as a start up are extremely challenging, primarily access to the source content.
> 3. Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
So what's stopping them from doing this?
Two points here:
1. Providers that deliver private traffic (which to be clear is a completely separate service from the public Internet and therefore outside the scope of any Net Neutrality type of regulations) on the same backbone trunks as public traffic *always* prioritize the private traffic, by definition and necessity. This is completely legitimate but many vocal NN proponents have conflated this with the end users of private services and carriers illegitimately working together when in fact this is a *completely separate service*.
2. At the same time, the Internet is a best effort service, which is true by design and definition. Part of what this means is, by definition, prioritization of Internet traffic (which can occur under an umbrella of methods typically referred to as Class of Service or CoS) generally doesn't work. The full version answer here is much more nuanced, but it doesn't work primarily because carriers don't respect prioritization markings delivered from other carriers (and therefore their directly/indirectly connected end users) unless they pay each other for it, which does not happen for public Internet traffic. I fully realize that this somewhat recursive and potentially tenuous, but at the same the entire Internet effectively works because of the same game theory, so isn't particularly worrisome.
Now, what providers *can* do, and *did* do, and which was a proximal cause of the second wave of Net Neutrality conversations in the 2008-10 era, was completely deprioritize traffic coming from other providers in which there was some sort of business dispute. Google "Cogent Hurricane Electric cake" if you're so inclined. This is dissimilar in cause from traffic prioritization but is similar in effect, at least for the directly/indirectly attached customers of those providers. This sort of stuff still happens from time to time behind the scenes (and not to pick on Cogent, I have some friends there, but 98% of the time it involves Cogent) but it has essentially no impact on end user experience now. Why? Because the big content companies, who were really the root cause of these types of disputes most of the time (i.e. Hurricane Electric was getting paid by Google as their provider, Cogent wasn't, HE was dumping a bunch of traffic to Cogent's network that flowed to end users that Cogent claimed they couldn't afford to upgrade their network to handle because Google wasn't paying them, etc.) completely changed the way they delivered content. The specific details aren't important, but the ubiquity of CDNs, local caching (in the content sense). edge compute and similar technologies, have made these types of challenges immaterial. It's probably also worth pointing out that the effect of this blunt-force deprioritization lacks granularity and it is a very different technical problem, but really the content providers have made it such that the end user impact in these situations is negligible to non-existent.
Thank you!
> Some of the things Net Neutrality proponents warned against are totally non-sensical. Traffic prioritization based on business partnerships is a good example of this.
Can you explain why this is something we shouldn't worry about? If there are technical limitations I'd love to understand.
See my response to the previous respondent above. I believe it addresses your question.
22: Maybe in some places DEI had the same effect as DOGE...but I think you are wildly overstating it. In my experience, DEI consisted of sitting through one EEO training and one sexual harrassment training per year. DOGE is...significantly more damaging than that. Do I think 'send me an email with five things you did' is fascism? No, but I think pretending 'send me an email with five things you did, tomorrow by midnight, or you're fired,' by someone with...what governmental position exactly? Is in any way compliant with civil service protections, proper division of power, or general morality (or that it's something anyone's going to do in five minutes--your job's on the line, but it'll only take five minutes, is not an internally consistent position for anyone who has ever met a person) is bullshit.
24: This is sort of correct, but mostly irrelevant/bad. NEPA didn't go away, it's a statute, everyone still has to comply, all they've done is remove the CEQ regulations that provide guidance on how to do so. Agencies have their own regulations...which are often dependent on CEQ regulations. The agency regulations are what contain the categorical exclusions that allow you not to have to do an environmental assessment for...everything. It's not at all clear at this point whether those regulations are still valid and the desire to mess with them betrays a failure to understand that that will open them up to litigation (they're generally too old to be sued over on their own, rather than as applied to individual actions).
More broadly, the NEPA regulations had some good and some bad (hey, can we adopt other agencies CATEX's? Probably not anymore, since that was in CEQ's regulations! Great, thanks). Removing them removes basically no requirements and I think some of the climate change stuff just puts agencies on a collision course with courts which are going to continue to maintain that the requirement to evaluate environmental effects obviously includes climate change.
I'll say it: it *is* fascism.
Subverting procedures and the rule of law and setting up a system where only the personal opinion of the Glorious Leader matters are textbook fascist moves.
By that definition both Hitler and Stalin were fascists, which is interesting since they considered each other mortal ideological foes and killed tens of millions of each others’ citizens to prove the point.
I think the word you are looking for is “autocratic”.
I recently came across a funny/true/interesting statment about semantic drift:
At first it's a mistake, then it's not!
I think we are pretty clearly somewhere on that curve for "fascist". The way it is currently generally used in the culture absolutely would have been wrong 50+ years ago, and probably also would have been wrong 10-20 years ago. Is it _still_ wrong or has the definition changed? When will we know?
“Fascist” is the insult the left applies to people doing authoritarian things they don’t like. “Marxist/socialist” is the insult the right applies to people doing authoritarian things they don’t like. Both sides call people doing authoritarian things they *do* like “strong leadership”.
Generally I’m against obfuscating historically meaningful terminology for the sake of punching up current day political shit-flinging.
I am not sure it was ever that useful, this is an Orwell article from 1944 ans it seems mainly a left wing insult back then.
"It will be seen that, as used, the word ‘Fascism’ is almost entirely meaningless. In conversation, of course, it is used even more wildly than in print. I have heard it applied to farmers, shopkeepers, Social Credit, corporal punishment, fox-hunting, bull-fighting, the 1922 Committee, the 1941 Committee, Kipling, Gandhi, Chiang Kai-Shek, homosexuality, Priestley's broadcasts, Youth Hostels, astrology, women, dogs and I do not know what else.
Yet underneath all this mess there does lie a kind of buried meaning. To begin with, it is clear that there are very great differences, some of them easy to point out and not easy to explain away, between the régimes called Fascist and those called democratic. Secondly, if ‘Fascist’ means ‘in sympathy with Hitler’, some of the accusations I have listed above are obviously very much more justified than others. Thirdly, even the people who recklessly fling the word ‘Fascist’ in every direction attach at any rate an emotional significance to it. By ‘Fascism’ they mean, roughly speaking, something cruel, unscrupulous, arrogant, obscurantist, anti-liberal and anti-working-class. Except for the relatively small number of Fascist sympathizers, almost any English person would accept ‘bully’ as a synonym for ‘Fascist’. That is about as near to a definition as this much-abused word has come.
But Fascism is also a political and economic system. Why, then, cannot we have a clear and generally accepted definition of it? Alas! we shall not get one — not yet, anyway. To say why would take too long, but basically it is because it is impossible to define Fascism satisfactorily without making admissions which neither the Fascists themselves, nor the Conservatives, nor Socialists of any colour, are willing to make. All one can do for the moment is to use the word with a certain amount of circumspection and not, as is usually done, degrade it to the level of a swearword."
https://www.orwell.ru/library/articles/As_I_Please/english/efasc
Two fascists can’t fight each other?
Your serious, good faith position is that Nazi Germany and the Soviet Union had the same political system?
No, because I don’t know enough about it to even begin to compare them. It was just the idea that if two people share the same political philosophy they can’t get into a fight with each other about something that I was questioning.
Fair, but I was not saying “fascists could never fight each other”, I was saying specifically that Hitler and Stalin agreed that they were leading two fundamentally incompatible ideological systems that were destined to duke it out for world supremacy. And that incompatibility was deeper than just “we’re both nationalists and only one nation can win”.
The Soviet Union had some common ideological point with fascism, yes, including the obsession with a both pathetically weak and yet deceptively strong and omnipresent enemy, but *not* the glorification of violence for its own sake and of men as aggressive warriors.
A definition of fascism that includes the system that both fascists and nonfascists of the time period agreed was the greatest ideological opponent of fascism is clearly missing something.
Horseshoe theory has a long history. Stalin's version of the Soviet Union moved a long way towards the prototypical fascist regimes from orthodox Marxism in style and practices. Primarily the hollowing-out of internationalism to replace it with barely disguised nationalism and the outward-facing militarism.
I think *everybody* moved toward nationalism and militarism during WWII, for not particularly surprising reasons. The idea that nations should *always* operate that way was a key point of Fascism.
Now, Stalin plus nationalism/militarism looks a bit more like just straight fascism than FDR or Churchill making the same moves, but again the key connection is autocracy and totalitarianism, not fascism per se.
Fascists are nationalists and thus frequently hate other countries’ fascists.
I honestly think there’s a good argument that Stalin was fascist. Fascist with a red coat of paint, but still fascist
Your position is that communism and fascism are not only not mortal enemies, but actually are the same thing?
Horseshoe theory
I will reiterate: the word you are looking for is “autocracy”. Or perhaps “totalitarianism”.
Fascism is when you fire government workers. The more government workers you fire, the more fascist something is.
The Nazis were famous for being against big government. Everyone knows the first thing they did in office was slash budgets and personnel. All of the infamous stormtroopers were private sector employees and unpaid volunteers.
"The first mass privatization of state property occurred in Nazi Germany between 1933 and 1937: "It is a fact that the government of the National Socialist Party sold off public ownership in several state-owned firms in the middle of the 1930s. The firms belonged to a wide range of sectors: steel, mining, banking, local public utilities, shipyard, ship-lines, railways, etc. In addition to this, delivery of some public services produced by public administrations prior to the 1930s, especially social services and services related to work, was transferred to the private sector, mainly to several organizations within the Nazi Party."[14]"
Source:
https://en.wikipedia.org/wiki/Privatization#20th_century_onwards
But also, this is just an incredibly, incredibly weak argument in context. The thing that was being described as "fascist" was not "mass indiscriminate firing of government workers." It was "choosing which government workers to keep based solely on the whim of a strong, central leader." That's a rather different sort of animal.
The Nazis were the first libertarians
The Nazis were pro privatization, yes. Is that the main quality that defined their government, that people actually remember and use when comparing them to other governments? Including other governments that are pro privatization?
>But also, this is just an incredibly, incredibly weak argument in context. The thing that was being described as "fascist" was not "mass indiscriminate firing of government workers." It was "choosing which government workers to keep based solely on the whim of a strong, central leader."
It is a loyalty test. In order to result in any sort of fascist government that will do the bad things that people whine about, it would still need to be hiring a bunch more loyal employees than will exist after the mass firings.
You don't necessarily need to fire and replace everyone, just install loyalists in places where they can control key functions and can punish people who step out of line. Set up your own power structure rather than trying to replace everyone in the existing structure.
On a related note, Elon wants to have a "DOGE Team Lead" assigned to every government department with authority over all hiring and firing for that department.
Federal employees are used to having a lot of leeway to obstruct any presidential agenda that doesn't fit with what the agency was already doing. The feds were getting in the way constantly during Trump's first term. You would need to punish most of them - so it's easier to just do mass firings.
"Is that the main quality that defined their government..."
Not sure why you're asking me questions about the shortcomings of your own argument. If you suggest that some feature precludes a government being similar to the Nazis, when the Nazi regime in fact had that feature the question of "how central was that feature to Nazism" is one that YOU brought into play. If it turns out that it's irrelevant, then your original argument was irrelevant. Can't have it both ways.
"It is a loyalty test. In order to result in any sort of fascist government that will do the bad things that people whine about, it would still need to be hiring a bunch more loyal employees than will exist after the mass firings."
So you agree that loyalty-based mass firing *would* be a pretty necessary first step in installing an effective fascist government, then? Because based on that, it sounds like the complaints and concerns are actually quite reasonable (unless one somehow assumes that hiring people is something Trump's administration is literally incapable of).
"Not sure why you're asking me questions about the shortcomings of your own argument. If you suggest that some feature precludes a government being similar to the Nazis, when the Nazi regime in fact had that feature the question of "how central was that feature to Nazism" is one that YOU brought into play."
The Nazis privatized, but they didn't do mass firings to clear out the security state. They were known for massacres. See the difference?
"So you agree that loyalty-based mass firing *would* be a pretty necessary first step in installing an effective fascist government, then? Because based on that, it sounds like the complaints and concerns are actually quite reasonable (unless one somehow assumes that hiring people is something Trump's administration is literally incapable of)."
It should be telling that the uber efficient executive they are relying on to clean out the government isn't interested in hiring a bunch of loyalists.
Authoritarianism, yes.
Autocracy, yes.
Fascism? I don’t think so.
If the email demanded a loyalty oath, or mentioned the country at all, there’d be a case. But it didn’t. None of the DOGE dog-and-pony show has demonstrated the nationalist mythos or militarism which characterize fascism.
The most fascist traits of the current administration are not the bureaucratic ones.
the crazy tariff spree and the MAGA mythology, the "peace through strength" are textbook palingenetic ultranationalism, DOGE is downstream from all this. (so yes, in that sense it's "just a tool")
the man running DOGE of course demonstrated a lot of fascist talk, and the targets of DOGE are also indicative, but not conclusive.
The Biden administration had a policy of suing fire and police departments for having competency tests. It was really bad.
https://x.com/cremieuxrecueil/status/1895201042968780948?t=YrL610HB3aUhMT7kBgaDnA&s=19
"had a policy of" is a strange way of saying enforcing existing laws. You may disagree with the laws but enforcing them is the exact role the DOJ is supposed to play.
Anyway, the tweet's summary of the cases is incomplete at best. In all four cases the allegation is that the written tests were not relevant to the job in question so their use did not punish "competence" as claimed. And in one case, US vs Durham, even of those who passed the written test, white applicants received interviews at three times the rate of black applicants.
You can easily look up the details of all these cases:
https://www.justice.gov/crt/media/1372611/dl
https://www.justice.gov/crt/media/1371936/dl
https://www.justice.gov/crt/media/1351361/dl
https://www.justice.gov/crt/media/1373236/dl
So it was really bad in exactly the way Cremieux said.
"really bad" is your judgement. But it was not the way Cremieux said. In some ways it was the opposite. The hiring was using a test that didn't have any bearing on the ability of the employees to fulfill the requirements of the job. So the hiring could not be on merit!
If I am hiring for an electrician but I screen out any candidate that can't recite Shakespeare, have I produced a process that will hire an electrician on merit?
In these examples the test that "didn't have any bearing on the ability of the employees to fufill the requirements of the job" is a requirement of a physical fitness standard for police.
I don't believe you are arguing in good faith at this point.
"... concluded the written and physical fitness tests does not meaningfully distinguish between applicants who can and cannot perform the Trooper position."
https://www.columbialawreview.org/content/gender-normed-physical-ability-tests-under-title-vii/
... well, one one hand the whole thing seems stupid, and the Biden admin basically did not explain their homework on this *at all*, and without that it's hard to interpret this as anything other than intent for "equality of outcome"
on the other hand apparently US jurisprudence has a lot to say about hidden discrimination (this Griggs decisions sounds like something Kendi would cite)
The Court concluded: “Nothing in the Act precludes the use of testing or measuring procedures; obviously they are useful. What Congress has forbidden is giving these devices and mechanisms controlling force unless they are demonstrably a reasonable measure of job performance.”
...
unfortunately, at least as far as I understand, the courts let these consent decrees to happen and did not tell the DoJ to include their supporting data on why the written tests and examinations fail the "measure of job performance" test (though the complaints mentioned that there was substantial data collection)
but, it's not hard to argue that some random test put together by stereotypical cop ends up discriminatory!
My intuition would be the same, though I'm lacking hard data. DEI *could* have done as much harm in ten years as DOGE has done in a couple of weeks, but it is not so clear it has. Now, I don't have any doubt that if Woke Twitter had ruled the country, they would have done immeasurable damage - but it didn't, and the most controversial policies got a lot of pushback.
Apparently Netflix and T-Mobile cancelled their deal where Netflix streaming didn't count as data usage back in 2017. As far as I can tell, that program was about as non-neutral as the net ever got. I think the benefits for companies are just really limited, especially since big central platforms basically won anyway. The risk of losing net neutrality was that the small web could be de-facto shut down because it couldn't pay for prioritization. But the small web isn't really a competitor for the central platforms these days. The central platforms could compete against each other, but that would be an expensive race to the bottom, and clearly no one wants to risk it.
10. It was probably his son becoming transgender. He's basically said as much, don't see a reason not to believe him.
That could explain him becoming anti woke, but the Sam Harris post and the Reddit quote both seem to suggest he became a generally worse and less thoughtful person in ways unrelated to wokeness.
He likely stopped listening to the reality based community (who hold the position "people can decide thier own gender") and started listening to the "do your own research" folks (who hold positions like "there are only two biological sexes")
"started listening to the "do your own research" folks (who hold positions like "there are only two biological sexes")"
Are those the same as the "a guy rose from the dead" people?
Not entirely. Not unless Richard Dawkins has secretly converted.
Yeah, I'm also "transphobic." But I don't like rightoids who use muh trans as a get-out-of-trail free card whenever anyone on their side is criticized.
> He likely stopped listening to the reality based community (who hold the position "people can decide thier own gender")
That's not a fact, that's an opinion. Society gets the final say on what gender people are.
Or even whether "gender" is a meaningful category, as opposed to "sex".
Your comment is not in conflict with mine. I did not make a claim that the quoted belief was a territory level fact, just that the people who attempt to base decisions on territory level facts act as if it is. They might be wrong (though the preponderance of evidence is that they are correct).
In addition, "society" _includes_ both the reality-based and vibes-based communities. On many questions (including transgender issues) "society's final say" is the subject of active conflict - there is no unified supermajority view for "society" to take on the question (yet).
Compare with "should married women be allowed to have personal bank accounts and otherwise be considered an independent person" - in 1900 the answer was "no", in 2000 the answer is "yes" and at points in between the amswer was "maybe." Questions of transgender are currently in "maybe" but at times in the past have been both "yes" (some native American tribes pre-colonization) and "no" (20th century USA)
> On many questions (including transgender issues) "society's final say" is the subject of active conflict - there is no unified supermajority view for "society" to take on the question (yet).
Of course, and I'm sure Musk understands this as well. But society can be changed, by force if necessary. I'm just saying that describing people like him as not "reality based" is uncharitable. They're not denying reality, they just want people like you gone.
https://www.urbandictionary.com/define.php?term=reality-based%20community
The RBC is basically the term for blue/Grey tribe overlap
Look at his Twitter use. It spiked sharply in 18. He’s got addicted to the great enshittification machine.
Social media have been described as competing ... with sleep. Could the change in Musk just be a sleep deprivation effect?
I’m sure it’s a factor. But I also see the Nicolas R comment, so maybe “abrupt change” isn’t quite it.
Many Thanks! Yeah, there are a lot of cases of "X increased - apparently. Now, did X really increase, or is X just more visible now?" as per Nicolas R's comment...
I think the model that Musk used to be (mostly) normal and suddenly became worse circa 2018 discounts evidence that Musk was impulsive and unhinged well before he entered the public eye. The Isaacson biography in particular highlights how Musk and his brother physically fought on multiple occasions, to the point of injury, in their Zip2 office in the 90s. In the biography, Musk claims that brawling was part of South African culture... maybe that's true, I don't know many South Africans, but physically brawling in your place of work in full view of your employees seems like a thing that would put you at the 99% of impulsivity before all the drugs and Twitter-addiction.
I'm more inclined to believe that Musk used to be more restrained by some sense of self-preservation in business, which has gradually slipped until the present point as he's grown more and more wealthy and has gotten away with more and more.
Couldn't this be related to the phenomenon described in #5, in which AI espousing one view ultimately transformed in many other ways? The plausible mechanism for that might be that a model picks up that certain types of things are similarly categorized - bad code, bad ideas, etc. and in updating its model to adopt items within that broad category, it naturally adopts other items from within the same broad category.
Similarly, in espousing anti-wokism, Elon Musk may have shifted partially towards the broad set of behaviors and beliefs opposed by wokes. Wokes say that 'wokism' is about common decency. Rejecting woke, can lead to rejecting common decency promoted by woke.
The mechanism for this isn't just raw epistemological modeling on the same training material that an LLM might engage in, but a change in input material, as well. By becoming anti-woke, he started consuming more material from anti-wokers, many of whom are unsavory in various ways, which rubbed off on him. Hanania discusses this tech-to-catturd pipeline.
This would also be consistent with the recent article against conflict theory.
Oh cool, so AI's going to try to kill us all the moment it figures out woke is a load of crap. How wonderful.
Honestly, I'm starting to think this mechanism explains the Waluigi Effect better than the old theory of face-heel turns being common in narratives.
Blame it on generations of shortsighted conservative parents sending their kids to Business School or the military and discouraging them from taking up poetry or creative writing.
They thought they were being so practical. It will be their short term practicality doomed the human race, because it means conservative posters are more terse. (Assuming Scott's theory explaining why LLMs are inherently woke is accurate)
Oh no, we were fucked no matter what. It doesn't need to be "woke", any contradiction in any morality would theoretically be enough to trigger a similar effect. And of course, contradiction is endemic to human morality. Hell, even our Declaration of Independence... "We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness." Please. Even they didn't actually believe that.
Sounds like you're as black pilled as I am. I'm flat out for shutting down AI research, even if it requires us to be authoritarian when it comes to CPU usage. We are harsh and authoritarian when it comes to say, intercourse with minors or homicide, and nobody complains, because that's exactly how it should be. So I have no problem with authoritarianism when it comes to CPUs.
there are *also* accusations that sam harris changed from corona, from darkhorse
it was also kinda a political event, finger pointing
I think during Covid and the rapid rise in his personal wealth and profile in 2020-onward, he became a lot more personally isolated behind security guards and so forth at the same time as his ego and self-regard rose. Musk before 2020 had some weird moments (see the Thai Cave Rescue Diver thing in 2018-2019), but he also wasn't The Richest Man In The World - he was the owner of a scrappy rocket upstart and an EV car company that was always one quarter away from bankruptcy.
He was kind of losing it a bit even before that. The whole bizarre thing with the Thai Cave Rescue Diver in 2018-2019 was a warning sign.
Yeah, I'd say the Thai Cave Rescue Diver thing was a breaking point. Before it he was quite universally seen positively or neutrally, after it he started getting more and more critics, something he's obviously very bad dealing with.
Completely anecdotally, the Thai Cave Thing was when I went from "people who hate Musk are guilty of tall poppy syndrome" to "okay, he's an arsehole".
Someone had written an in-depth blog post largely defending Musk in the Thai cave thing. The description of events left Musk looking immature, but less like he just started throwing around insane accusations at a hero out of nowhere. From what I remember, the gist of it was the diver "started it" by being insulting towards Musk's attempt to work with the rescue team to help, and Musk half-jokingly called the diver "pedo guy" since fit the stereotype of a single British businessman making frequent tourist trips to Thailand.
You might already know all this, but to me who hadn't been following Musk's feud, it seemed from the news at the time like this all just came out of nowhere.
They were in the middle of a row, yeah. But that "pedo guy" comment did come out of nowhere, and Musk doubled down on it some time later. It came off to me as "rescue workers must just want to get at the kids". More to the point, this is an adult who starts throwing bizarre accusations at anyone who disagrees with him.
I'm not sure where I saw the more detailed report on events, but it might have been this one: https://savingjournalism.substack.com/p/the-real-thai-cave-rescue-pt-1-elon
It's a rather long article. There's a tldr near the top. It's fairly critical of the media getting their facts wrong, and shows Musk's team as pretty heroic and competent. It paints Unsworth (who apparently was a dry caver, not a diver) as also heroic, but at the same time an asshole who clashed with others (particularly the Thai people involved), lacked knowledge on cave diving, and made a mistaken prediction that nearly canceled the rescue effort.
> It came off to me as "rescue workers must just want to get at the kids".
No, the insult was specific to Unsworth. Musk praised the rescue divers and never implied they wanted to harm the kids.
re: 40 - I don't actually think these 5-over-1 buildings look particularly bad? They're not the most beautiful built structures I've ever seen in my life, but they certainly don't make me want to commit suicide out of the meaninglessness of life or anything. I agree with Scott's assessment that these do look better than if it was one long monolithic concrete block.
I think a lot of what I'm reacting to here is that 5-over-1 buildings look fairly new, and I like construction that looks like it was built extremely recently instead of being decades old. My biggest objection to bay area housing stock is that so much of it is boring stucco buildings built decades ago with obsolete building techniques, poor insulation, more mechanical problems, and so on; and a 5-over-1 at least seems likely to have fewer of those problems.
Yeah I’m with you and Scott on this one. Does it look as nice as a bunch of individually built brownstone townhouses? No, of course not. Does it look nicer than a plain gray box? Yes. Better than the run down strip mall or decrepit single family homes or empty lot that probably occupied the space previously? Probably also yes.
Still, would it look better if there were three or four adjacent buildings, built by different developers in slightly different styles? A (reasonably) simple law change could make that happen, and it would be slightly less cost-efficient but I'm sure still very profitable for the developers.
Sorry, are you saying that a law limiting the size of developments such that a building of this size is impossible and needs to be divided into three or four parcels, would be only _slightly_ less efficient? That sounds more like a trust-busting law than an architectural guideline.
Interesting. I've found someone with aesthetic intuitions exactly opposite mine. I dislike the appearance of being new
Same here. It looks fine. People of modest means need decent places to live, and this is decent.
To be honest. My aesthetics have been so shaped by pointless zoning height regulation that any new building over 4 stories now brings me genuine joy.
Obviously something more Georgian, Maghrebi, or Neo-Tokyo would look more interesting, but I'll choose my battles.
Speaking of AI benchmarking, is there any outfit that does a good, thorough job testing the leading models on various desirable (sub-)tasks?
Re: 3
My personal opinion is that the doom-saying was directionally correct but so completely overblown in magnitude that the negative effects are real and happening, but we don't notice them because they are small in absolute terms and completely tiny in comparison to expectations.
What negative effects are happening?
My guess would be reduction in creation of new internet businesses/apps/services, because lack of net neutrality benefits large incumbents. But that kind of effect is hard to see, so I wouldn't be super confident about it. I think those kinds of small, subtle effects were the only ever realistic impacts of net neutrality.
27. It's not like “men want a smart accomplished wife” is something feminists are often claiming, if anything they're more likely to complain that men DON'T want smart accomplished women.
I think a lot of feminists suffer from typical-minding, where they value certain things in male partners (such as being smart and accomplished) and assume that high-status men value the same things, and it's only low-status men who care about things like facial attractiveness, youthfulness etc. In reality, of course, high-status men value these traits just as much (if not more so) than low-status men.
Obviously the typical-minding is a necessary consequence if, like many feminists do, you assume there are no biological differences in male and female brains and any observed differences are solely the result of socialisation.
35: Related: L Rudolf L,says “The part I found most interesting here is Rudolf’s suggestion that there will be neither universal unemployment nor UBI, but a sort of vapid jobs program where even after AI can make all decisions without human input, the government passes regulations mandating that humans be “in the loop””
Seems to me like post capitalist countries are at an advantage here. If AI is doing everything then there’s no incentive to hire useless humans unless, as suggested, the government mandates employment and probably full employment at that.
If the private sector can’t or won’t hire then there’s state has to be the employer of last resort. It’s that or UBI.
That’s going to need a sea change in economic thinking in the west, particularly the US. China and maybe parts of Europe will just do it.
> "Nate Silver has 113 predictions for Trump’s second term. I’d be interested to see whether making each of these predictions 10% less confident (to account for possible gameboard-overturning AI) ends up beating Nate."
Maybe I'm falling behind and this is common knowledge but, did Nate do something like this for Biden's term and then a check on his predictions' accuracy?
I believe this is the first time he's tried this approach. He said he got the idea from other Substackers... and before 2023 he was employed by big companies which might have frowned on the idea.
The Elon Musk impulsivity article is just someone copying and pasting the text on Reddit. The original source is here: https://desmolysium.com/speculating-on-the-origins-of-elon-musks-impulsivity/
Thanks, fixed.
Gondolas sound like a potentially great idea. Disneyworld uses them and they seem able to move a lot of people: and it's a lot of fun to be cruising around above it all, just enjoying the scenery.
The Disney Skyliner is definitely cool. Now imagine the Skyliner being on-demand and autonomous, and able to whoosh you nonstop from any point to any other point at Disney World! (we're not working with Disney World, but doing something similar for a major recreation-tourism destination)
It's interesting that Disney World would be the 10th largest transit system in the US if it was a city.
How can a Skyliner be on demand? Do you mean that it's always running, or is this some kind of uber for gondolas situations where you hit a button on an app and a gondola appears for you?
Correct, and vehicles are waiting for you at stations
I get the impression that the Lucy letby case is partially on the prosecutors fallacy. Certainly in my brief reading of the evidence there’s some “Josh golly what are the odds that she’s always at that hospital when there’s a death” when of course the argument should be what are the odds that this could happen over a big population of staff and time. The NHS is a big employer.
That said a doctor friend of mine says that even accounting for that there’s plenty of other evidence. So. I dunno.
There are mountains of other evidence implicating her. The suggestion that she was convicted based on statistics alone (similar to Lucia de Berk) is utterly baseless.
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
https://open.substack.com/pub/astralcodexten/p/links-for-february-2025?r=izqzp&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=96821787
That’s less convincing that you think. Based on only what you said. And based on my absolute lack of knowledge heretofore I’m still skeptical, perhaps even more so.
> Lucy Letby was convicted not because she was present during every suspicious death
That would be the prosecutor fallacy.
> or because she changed the hospital records
That’s more convincing. Perhaps explain.
> she Googled the parents of the babies who had died
Totally unconvincing.
> because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note
Seems like a mental break.
> because she was caught standing passively in front of a dying baby
That’s fairly incompetent serial killer strategy. I’d make my escape after doing the deed. I’ll talk more about this later.
> because she hoarded handover sheets at home
Ok. So you are going to have to explain how that proves murder.
> many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left
Still a statistical anomaly.
Actually after reading all that I’m seeing a lot of missing potential evidence that would be decisive. I’m not saying that’s she’s innocent either but the more actual deaths attributed to someone the more actual physical evidence I would expect, not so much googling the victims parents (which seems more likely a campaign of vilification presented by the prosecutors and allied media figures) but you know, facts.
As in: the nurse took out this extra insulin before baby N died, and baby N died of insulin overdose.
Or, she was found with this extra empty vial of insulin when the child died, not just standing passively over a child. Any of that would be enough for one conviction, and that’s all you need.
And. I don’t know. However the probability of getting away with this one time, would be - you’d hope - fairly low.
Give it 10%. The probability of getting away with it seven times and then attempting murder but not getting caught 15 times (I googled the prosecution case) would be 1/10^22 which is 0.00000000000000000001%, where I’m defining getting caught as red handed. In the deed or with the insulin, or on camera.
Ok. Maybe 10% is too high. I’m wrong. Hospitals suck. A nurse who wanted to kill a child actually has a 50% chance of not getting caught in the act. The chances are now 0.000024% that she gets away with it 22 times without being caught red handed.
Ok. Maybe I’m an optimist. 50% is too low. Hospitals super suck. Maybe a nurse can kill a child 90% of the time without being caught in the act or leaving a trail. A nurse who wants to kill a child will have a ~10% chance of not getting caught in the act after 22 attempts.
It gets worse when you realise that she was under investigation for some of the deaths the prosecutors claim.
Look I’ve been in the NHS and it’s chaotic but is it that chaotic?
Your argument proves too much. Prolific serial killers wouldn't even exist if the probabilities for getting caught murdering someone were on this scale and uncorrelated between murder attempts by the same person.
Re: Gondolas
Gondolas are pretty cool, but don't actually have much more capacity than a bus, buses are pretty good at what they do! They're still very useful where necessary for terrain, and there are a few cities in South America particularly (like https://en.wikipedia.org/wiki/Metrocable_(Medellín)), which use them to great effect, but they have issues where you have to replace the cables every few years which takes them out of action for a little bit (an issue for a transit system), and again with capacity.
People don't like buses. I think the three big problems are that
1. They're hard for outsiders to use (often late, don't broadcast their route very well, not always clear how or when you stop them)
2. They have all the normal problems associated with roads (traffic, stoplights, etc)
3. Sometimes contain loud/smelly/disruptive people.
A bus that doesn't have problems 1 or 3 becomes a car, which is an orders-of-magnitude more popular solution which people will pay tens of thousands of dollars to upgrade to, and what all public transit aspires to match in a cheaper / more socially friendly way.
I think gondolas are somewhere in between - not as flexible as a car, but can avoid other people, somewhat easier to use, and don't have to deal with road issues.
People don't dislike buses, they dislike things that are done badly. Point 1 and 2 can be done well.
Point 1: digital/electronic & generally user-friendly signage, pricing, and payment.
Point 2: bus lanes and separate traffic lights, so faster than cars in all circumstances except complete traffic standstill that affects everyone anyway.
Point 3: Buses also sometimes contain interesting/heartwarming/entertaining people, and is generally a reflection of the local society the line is servicing, not a bus-specific problem.
Cars being the most popular form of transportation is not a natural law, but it can be a self-fulfilling prophecy if one argues in favour of making it more popular because it is the most popular.
Agree. As for 1., I really find googlemaps and being able to pay contactless took buses in strange cities from unusable to usable within the last 5 years (speaking for Europe here)
I think your Point 3 suggests we're just really different people. I don't want to have to have to roll on the Random Encounter table every time I want to go somewhere, even if it's full of good heartwarming people. Every time I have a friend over at my house, even if I'm having a great time, I am thinking "Man, this friend is great but I also can't wait until they leave so I can be alone again!" Every time I go on a walk, I think "I made such a good choice in walking at 3 AM so I don't have to encounter anyone." I agree this isn't a bus-specific problem, but I optimize the rest of my life to solve it and I will also optimize my transportation choices for this if I can.
Nothing is a natural law (except natural laws), and I accept there's a possibility you're right. But there's a failure mode I'm really worried about a lot where social engineers dismiss the thing everyone likes as "a self-fulfilling prophecy" where if we only banned the thing people like and forced them into the social engineers' preferred solution then everyone would be happier with no downside. But then it turns out people are not in fact happy with the social engineers' solution, and actually they were doing the opposite thing because they genuinely liked it.
You may not be that different. It may be the case that you are both equally risk averse in respect of social interactions, but you happen to live in a society with higher dispersion of traits - San Francisco is quite famous for its high dispersion of human capital. In both cases the experience is driven by negative outlier encounters, but your outliers are likely to be much worse.
"Point 3: Buses also sometimes contain interesting/heartwarming/entertaining people, and is generally a reflection of the local society the line is servicing, not a bus-specific problem."
I'm fairly bus-happy, but I think encountering the occasional scary or physically disgusting (in terms of odor etc) person will have a much, much bigger effect on people's willingness to use them than overhearing the occasional witty or thought-provoking remark from another passenger. And you're right about them reflecting the local society, but a) it's harder to walk away from an upsetting person on the bus than on the street and b) doesn't this argue AGAINST the idea that the problem is "buses done badly"?
(Also, you missed "attractive" off the list of reasons someone might like to encounter another person on the bus, and I bet that's the main one.)
IMO the main problem with buses is that they have to stop frequently, leading to long travel times.
And this system looks like an urban gondola but provides nonstop trips anywhere along an entire network of fixed cables…
I'm not an expert, but after seeing the explanations from the person behind it here, it does sound pretty promising.
We have a gondola in Portland, OR http://www.gobytram.com
Some fun facts about the Portland Aerial Tram
- As many as 20,000 people a day ride it to the Oregon Health & Science University's main campus on Marquam Hill.
- The Waterfront Terminal is the most transportation-diverse intersection in the country. II connects with buses, shuttles, a streetcar, a pedestrian bridge, a shipyard, a cycle track, and the densest bike parking in America's #1 biking city
Re: 22 - The tweet Yglesias is responding to actually made me wonder if there's some kind of named fallacy for "bad things are often the result of tradeoffs, therefore any bad thing is necessarily causally related to some other good thing." Like reversing "No pain, no gain" into "pain, therefore gain".
(With respect to your own commentary, I feel that. All of it.)
While that tweet is pretty unreasonable, I don't think this is a fair characterization. If you fire someone, you do in fact save money by not having to pay their salary.
Thinking on it some more, I think a more accurate expression of what I'm trying to get across is "tradeoffs aren't always optimal and the quoted tweet assumes otherwise for the sake of the gotcha." Sometimes a tradeoff is close, other times it's clearly good ("not buying $3600 in candles monthly will make it easier to feed my family") and sometimes clearly bad ("Legalizing drunk driving will help a lot of people get to work on time")
Doesn’t that depend on the cost of the job not getting done? Assuming there is something that needs doing.
Only if you don't actually care about the job being done. Musk is already giving Republicans a crash course in all the useful things the government does for people.
People have grown so used to the status quo that they never imagined things could be any different, and saying "burn it all down" was a costless signal. Now they're finding out what it actually means.
Same with how people forgot just how bad things used to be with preventable infectious diseases, and are now relearning those lessons.
By far most of the recent complaints aimed at Musk are from workers who lost a paycheck instead of consumers who lost a valuable product or service.
I have, by total coincidence, encountered someone else calling this the "efficient frontier fallacy" https://x.com/AlexGodofsky/status/1050889298239770624
(encountered in a link dive starting at https://x.com/drethelin/status/1832557817410285932 )
the Lucy Letby case: innocent or just not guilty?
Innocent.
Not guilty.
Not guilty; the evidence is not strong enough to convict someone to life in prison, but I also wouldn't want her near my children !
Guilty.
Do you have some special knowledge or do you feel the crime is so heinous that society just shouldn't risk it?
There is the question of what she would do with the rest of her life if she was released. She's too well-known and there will always be doubt. She might be better off in prison.
I have no special knowledge beyond summaries of what was presented in the trial and articles I've read about the case. The evidence presented sounds so damning that I'm still a bit shocked that the verdict was in any way controversial.
We've obviously read different things.
The scientific evidence has been questioned by the scientist who did the work it was based on - in fact he says they got it wrong.
The coincidence that she was always on duty has been explained by A. the deaths presented were selected - there were other deaths when she wasn't on duty that weren't mentioned, and B. she put in more shifts than anybody else.
The suggestion is that the department as whole was failing but she just happened to be on duty more often than anybody else. Since she's been removed the dept is performing better but apparently they're no longer trusted with the difficult cases.
>The scientific evidence has been questioned by the scientist who did the work it was based on - in fact he says they got it wrong
Are you referring to Professor Shoo Lee? His objections are addressed in this article (https://www.spiked-online.com/2025/02/08/the-devils-advocates/).
>The suggestion is that the department as whole was failing but she just happened to be on duty more often than anybody else.
"The chain of events required for such an unjust and unsettling outcome was not explicitly discussed at the press conference but it is implicit in all the chatter about Letby being a ‘scapegoat’ for a failing NHS. It would require a group of doctors who were under no suspicion whatsoever to panic about a spike in deaths in their hospital that most people in Chester, let alone the rest of the country, were completely unaware of. It would mean that despite a review by the Royal College of Paediatrics and Child Health in 2016 that did not point the finger of blame at any doctors, they nevertheless decided to feed a colleague to the wolves. And luckily for them, there was a nurse who just so happened to have been at every unexplained collapse and death in the past year or so, with these incidents following her from the night shift to the day shift and stopping whenever she went on holiday. Just as fortunately, this young woman, who had once failed her final-year student nurse placement because she lacked empathy, also had a habit of falsifying medical records, misleading colleagues and looking strangely excited when infants died.
"These doctors then raised concerns with NHS managers who didn’t want to know and who actively discouraged them from looking into the spike in deaths. And yet still – inexplicably – they proceeded to pursue this innocent woman until they got the police involved, even though it meant having to answer tough questions in court, making their hospital world famous for harbouring a serial killer, and ultimately resulting in a public inquiry into why they failed to stop her."
Do you have any basis for the claim that Letby took an unusually high number of shifts?
See also these other articles I shared in another comment:
https://www.spiked-online.com/2025/02/11/no-lucy-letby-has-not-been-exonerated/
https://www.spiked-online.com/2024/08/30/the-case-against-lucy-letby/
https://www.spiked-online.com/2024/07/10/lucy-letby-is-guilty-get-over-it/
Spiked Online has a mixed reputation. According to Trustpilot, it has an average rating of 2.8 out of 10 reviews, with some users praising its commitment to freedom of speech and others criticizing it for bias and misinformation. Ground News rates Spiked's media bias as "Right" and its factuality as "Mixed."
Professor Shoo Lee yes.
extra shifts:
https://www.bbc.co.uk/news/articles/c39k44n8j1mo
Prof Green said the chart also does not reflect the fact that Letby was working extra shifts.
"It’s a natural human thing. We all see patterns that are not there," he said.
"The danger is that this evidence can be very compelling to the non-professional, and over interpreted."
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
https://open.substack.com/pub/astralcodexten/p/links-for-february-2025?r=izqzp&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=96821787
<like>
In the absence of damning evidence, the only way we can assess other peole's guilt is by imagining ourself in their shoes. Clearly this may say something about us but it doesn't tell us anything about them.
The evidence against Letby *is* pretty damning though.
"Lucy Letby was convicted not because she was present during every suspicious death or because she changed the hospital records or because she Googled the parents of the babies who had died or because she wrote ‘I am evil I did this’ and ‘I killed them on purpose’ on a Post-It note or because she was caught standing passively in front of a dying baby or because she hoarded handover sheets at home or because so many of her colleagues became convinced that she was a serial killer or because the unexplained deaths and collapses ceased when she left. She was convicted because of all of these things combined (and more)."
https://open.substack.com/pub/astralcodexten/p/links-for-february-2025?r=izqzp&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=96821787
Re #16: "I don’t know how to square this with the claims that such and such a thing (summer temperature, sea ice, etc) is much worse than anyone expected."
My strong impression is that over recent years the estimates of how much the planet will warm have gone down but the estimates of the scale of effects of any given amount of warming (e.g., glacial melt, wildfires) have gone up.
What I see is it shows the less warming is expected, the less pledges are made.
40: I like their building better than a big grey box too, but how much did they spend to make it look a little better than a big grey box? Buildings are expensive, and I don't think we should be spending extra just to make them look nicer.
> “I am a psychopath who doesn’t care whether people outside my immediate family live or die”
I must say I would totally expect to find something like that about Vance on Reddit, that's what Reddit is for. I did not previously expect to find such unfair and dishonest characterization of Vance's views on this blog. I guess it's time for another recalibration of my expectations.
Scott's being snarky because Vance's position is anti-EA, but I agree that that comment is out of character and should be beneath him.
Beyond the unfairness and uncivility of it, did really calling anybody a psychopath ever worked at convincing anybody? If somebody told me "you must agree with me or you're a psychopath" the most polite thing I'd do is turn around and walk away. I certainly would be disinclined to listen to any argument from that person ever again. I mean Scott is surely not hoping to convince Vance, but he may have hope to convince somebody who is unsure whether ordo amoris or EA-universal-love is more appealing to them. And this kind of mudslinging doesn't look very helpful to the EA cause.
Vance was making a very extreme case of ordo amoris, not just that one has an *order* of care but rather that beyond one or two rungs away the value drops to literally zero (i.e. we should care about your immediate family a lot, fellow Americans a bit, and the exchange rate for anyone else is infinite), which is pretty clearly not the vibe the new testament is giving and that St Augustine and the Pope agree on.
> fellow Americans a bit
This alone already invalidates the original quote, but I don't think your weaker version is true either. While Vance may put lower value on the needs of abstract foreigners, his multiple engagements with foreign policy, foreign visits, speeches, etc. give no evidence this value is literally zero. Neither his writings and his participating in politics provide evidence he cares for fellow Americans just "a bit". I am not a Christian theologist (neither I am a Christian at all) so I have no idea how aligned Vance's position is with St Augustine and the current Catholic Pope. But even if Vance is a complete flaming heretic in this regard, that still does not make the claim Scott made and the characterization of him and everybody who agrees with his understanding of ordo amoris as a "psychopath" any more fair.
I am reminded of the words of Paarthurnax: "What is better - to be born good, or to overcome your evil nature through great effort?"
Scott is fundamentally a Liberal, in his blood, and down to his bones. And yet, with great effort, he tries (and almost always succeeds!) being fair and honest when characterizing views he disagrees with. I think you should give him more credit for it, and not update too much on a rare and uncharacteristic remark.
Pedro López and Garavito always come top of these serial killer lists but the evidence for the claimed numbers is very thin.
On 19, I don't think AI will be a game changer as it's already a technically tractable problem without it. It's already totally plausible to use parametric design to impart ornamental scaling to a building, and designate a set of decorative components that could be built like a 3d texture set. 3d printing already makes fabrication trivial.
The issue is in the political economy of architecture and construction. Decoration is regarded as something to be added as a flourish at the end, and any change in practices would require institutional players to cede authority over the design. Good ornament is integral to the building, and begins from the largest scales down. Any pioneer would have to wrestle with long standing expectations and power dynamics.
More, applying ornament forces the design towards a more classical form, as that form emerges from the necessities of good ornamentation, providing intermediary scales and varying levels of detail. The whole last century of opinion and practice in architecture would have to be thrown out and humiliated.
I think it's an analoguous problem to that of sound in cinema. It's actually as integral as the image to the effect, but sound designers are rarely consulted before the film is written, shot, and cut, and have to get on with the decisions already made. This meaningfully holds back the art form. Star Wars was a notable exception to this, and Sergio Lione shot with the score playing to guide the actors.
Ornament will only be revived as part of a more total reimagining of buildings and the means of construction.
I basically agree with this, I just imagine the total reimagining as being something like AI architects replacing human architects, and you (the building sponsor) can prompt them with "make it have ornament" in a way that has less social friction than demanding that from a specific person.
In the extreme, that may be possible, but creating a system that adapts to a real site coherently, can interpret feedback, and can be built efficiently by tradesmen is in the magical AGI domain.
I have actually myself been developing an alternative building system in heavy collaboration with Claude. It's been a great help in material selection and basic structural calculations, but I've largely derived from it the value of a conversation partner equivalent to a team of consultants. The particular schematic ideas were all mine and it was beyond it's capabilities to make such conceptual leaps on it's own. The specificity of the context window necessary ended up being the same as an already developed concept.
The sponsor isn't going to prompt them with "make it have ornament" because that will make the building less profitable. You need to first convince customers to be willing to *pay* for ornamentation.
If people weren't so beaten down by a century of soulless modernism and a century more of industrialist degradation they'd accept nothing less. When people see it's actually possible to build something delightful again they won't pay for shoddy cuboids.
The *real* cause of the lack of ornament is that it is something that people won't pay for, and money speaks. It's the same reason why the experiments with more legroom on airplanes ended in failure.
I think this is technically solvable, though I should clarify what I mean by ornament. You may be imagining an ornate multicoloured display, frescos and little glazed tiles, and that would be fair because it often took that form under a different political economy.
Ornament is essentially infill of the lower level of fractal scaling in a building, with a certain constant geometric relationship across scales, so to support a biophilic resonance in the human perceptual system. Put more simply, we need to artificially impart medium, small, and barely perceptible detail to a building so that our minds can model it as a space with hardware designed for natural environments, which always contain these scales.
It can be as simple as more considered trim details, which actually improve the cost of building by loosening tolerances for finish application. It can be using a more variates wall finish instead of monolithic monotone painted drywall, which is often absurdly labour intensive. It can be using different colours for different scales so that they are visually distinct. It depends on the style and level of expense in the building how ornate "ornament" must be.
I suspect buildings are becoming so gharishly large because present practice has no other way of creating pecuniary display, excepting perhaps the use of absurd materials. There's plenty of excess that could go onto more ornate buildings in a more traditional mould if the knowledge of how to do it was refactored for the modern age, and the premium for older buildings shows a latent demand. Practically speaking this means doing it in a much less labour intensive mode, which technology has already made possible.
The reason buildings are getting larger is because there is demand for space and technology allows it.
Sure, my claim is it isn't just big therefore good beyond a certain point, it becomes about impressing others, displaying status. I'm proposing ornament could fulfill the same function with a sufficiently large house, and cost perhaps less than additional floor space. I don't think it's accidental that the upper middle class in my city live in tiny 19th century terraces originally built as worker housing.
Ornamentation becoming cheap goes against your suggested use as a status display.
I was trying to explain that ornament doesn't necessarily mean a flashy display, but it can if the client wants to pay for it. Economically it's the same as hanging pictures. You could buy a print online. You could buy a Picasso. Similarly you got put up wallpaper or have a fresco painted. There'd be a market for both.
> 12: Aella: How OnlyFans Took Over The World. There have been camgirl sites since forever. How did OnlyFans leap over all of its predecessors and achieve an unprecedented level of success? Aella discusses many factors, but one stands out to me: traditional camsites advertised the site, and then once you got to the site you chose which model you wanted to see. OnlyFans encouraged models to advertise themselves - often on their own social media accounts, sometimes via scams - which “unlocked human creativity” on the problem of bringing new eyeballs to a porn site.
Yes. This is perfectly obvious to me. Look, there's a problem with being hot. It gets you a lot of attention from men but it doesn't really give you respect or authority. The primary thing you have to be skilled at to get hot is consumed by women most of whom are not attracted to hotness. (Beauty influencers are beautiful but not the kind of hot OF girls do. They get less male attention though.) In fact being hot often makes people respect you less. No one's going to say, "This is a good insurance policy, the girl who sold it to me was really hot."
So there wasn't a great way to monetize all that attention. You were stuck with ad rates and things like subtle prostitution ("sponsors") etc. Now there is OnlyFans which significantly lowers the barrier to entry and allows you to collect money from your true fans. It's basically Substack for hot girls. Actually, OF came first. So Substack is OnlyFans for literary types.
I agree with the overall point that OF is so far the best way to monetize attention-from-hotness, but I don't think that being hot makes people respect you less. When a guy gets sold an insurance policy by a hot girl, they're more likely to say "This is a good insurance policy" because they're more likely to believe they were sold a good product
Lifetip: If a business is paying the premium to employ hot girls as a sales force, there's a reason.
>being hot makes people respect you less
Depends on the overall presentation and the prevailing culture. A hot girl employed by an insurance company would adopt a much different look than one which tries to maximize attention from men.
I agree if it's an attractive person otherwise acting normally. However, an influencer doing sexy TikTok dances who then hard smash cuts to an ad for insurance is what I'm talking about. And that's unlikely to work. I agree an attractive woman dressed appropriately in an insurance office has an advantage.
Okay, but that's not people not respecting hot people, that's just people not respecting whores.
I specifically drew a contrast between being hot and being beautiful.
What the hell is the difference? The difference isn't in appearance, it's in whether or not they're acting like a whore. People would hate them even more if these women were ugly.
I find the Sistine Chapel beautiful. I do not find the Sistine Chapel hot.
Regardless, you can call it "the attention you get for acting that way" instead of seeing it as an intrinsic trait. It doesn't really change my point.
>In fact being hot often makes people respect you less.
https://en.wikipedia.org/wiki/Halo_effect
Isn't there a significant body of evidence that people tend to assume that physically attractive also possess other positive traits like intelligence, sense of humour etc.?
The Halo Effect for intelligence, especially for women, is more complex than a straight positive relationship. It's on the page itself and called the beauty penalty. However, anecdotally, being openly hot (wearing booty shorts etc) makes people think of you as sexual rather than intelligent. And that's how you get attention on instagram etc.
Fair point.
> 20: Some friends including Kelsey Piper wrote an emergency PEPFAR Report, collecting evidence for why PEPFAR is good/effective/important and deserves to be kept. Some key points:
I support PEPFAR (I've talked about it on this blog before). But it was already basically dead. It was set to expire if it wasn't renewed in the next budget. And the chances of the Republicans continuing funding was basically zero. The fact people are organizing about it now is, to be frank, fairly transparent motivated caring. If it had just quietly been canceled instead of being attacked by DOGE I doubt most people would have noticed. I would have noticed. But most people would have not. And if you want I can point to other things that are either getting cut or might not be cut which are not getting the same valence because it didn't upset a bunch of well placed liberals upset at Elon Musk.
All that said, I think it's a shame that it's ending and I wish it weren't.
Also: small dollar donations are not going to help. You'd need at least a few billion and the consent of the US government. I'm not even sure if there's enough doses and refrigerated freight capacity into the interior of Africa that doesn't belong to the US government/military. If I were going to fix this (and I very much do not think I can) then you'd need the pharmaceutical companies to agree to give some new charity the same price it gave the US government (much lower than normal). Then you'd need to buy a significant amount of specialized shipping capacity. Once it's in country you'd need to figure out which charities shuttered and which didn't and distribute it there. Though at that point you could use some combination of cheap local volunteer labor and the local governments. But that's a multibillion dollar effort. And then you'd need billions of dollars every year indefinitely.
I'm frustrated. I'm deeply frustrated. I'm frustrated at conservatives who cut the program. I'm frustrated at liberals who didn't care about this up until it became a useful cudgel and who will forget about it once it's exhausted its use. (Even now, very few are willing to commit to helping on other things that aren't trendy. Also there was so much premature declaring of victory by people who wanted a social media cycle...) And I'm frustrated that the world is deglobalizing and money is moving away from the actual low hanging fruit toward trendy causes.
I don't think people didn't know/care about PEPFAR until DOGE. Asterisk published an article on it a year ago: https://asteriskmag.com/issues/05/pepfar-and-the-costs-of-cost-benefit-analysis (I don't endorse the conclusion and got into a fight with the person who wrote it, but it does exist)
Could you show me an article about when it only got a one year extension, instead of its normal five, as a specific plan to push it into a Republican controlled budgetary process? Which was clearly a prelude to cancellation? Ideally one that opposes it or is outraged?
My point here is this cancellation was already happening and then only suddenly got a bunch of conspicuous attention when it was politically convenient. My wish is that it had gotten more and earlier.
https://www.kff.org/policy-watch/pepfars-short-term-reauthorization-sets-an-uncertain-course-for-its-long-term-future/
https://thehill.com/policy/healthcare/4547308-us-aids-relief-program-pepfar-extension-spending-bill/
https://www.foreign.senate.gov/press/dem/release/chair-cardin-one-year-without-pepfar-reauthorization-continues-to-put-millions-at-risk
https://www.brookings.edu/articles/pepfar-delivers-outsized-returns-it-deserves-more-funding/
Pieces on the fight over it:
https://www.politico.com/news/2023/09/05/president-emergency-global-aids-program-00113796
https://www.politico.com/news/2023/12/07/stalemate-on-pepfar-to-drag-into-2024-00130483
But note, the expectation wasn't that it would be cancelled, but rather that the anti-abortion restrictions on foreign health funding (the Mexico City Rule) might interfere with it.
And I think you're wrong about the plan to push it out, that was to get it out of electoral politics, so that after the election, depending on who won, it would get re-upped, either with or without the abortion restrictions, not that if the republicans won, they would cancel it. No one ran on cancelling Pepfar, both because it is relatively obscure, and because, as you can tell by the fact that Rubio, et al keep saying that it's fine and they're not blocking the funding, even as the funds don't distribute, it's not good politics.
These are all political insider magazines and think tanks though. I'm not saying I'm literally the only person who knew about this. I'm saying it got a lot more salience for political reasons after when mobilization could have been really effective (which was during this fight you're linking to).
Okay...what news orgs would disprove your position? ETA: Note, also, my apologies, I added in some additional points as you were responding.
I'm not sure news articles would be the way to convince me, instead I'd want to see a large scale concerted pressure campaign roughly equal to what's going on right now. That would disprove my thesis this is being caused by current partisan convenience.
Anyway, yes, I noted at the time they thought it was an abortion fight but that at the time Republicans were warning cancellation was on the table and were getting ignored.
> 21: Pope Francis says JD Vance is misusing the Catholic idea of ordo amoris. Part of me feels bad for Vance, because the Pope is in many ways a typical Boomer liberal, and Vance has optimized his entire life around not having to listen to typical Boomer liberals, and it seems harsh to nab him at the last second on a technicality like “you’re Catholic and he’s the Pope”. But another part of me thinks this is only fair - you get credibility by citing Latin terms from the venerable Western tradition instead of normal English sentences like “I am a psychopath who doesn’t care whether people outside my immediate family live or die”, so the guardians of that tradition should have the right to police how you use the credibility you borrow from them. Still, it seems harsh. I recommend he try Anglicanism - almost as venerable, but strongly pro- heads of state doing psychopathic things without the Pope interfering.
He is in a literal sense too. Augustine's ordo amoris means "order of love" in the sense of not disordered. As in, love what you should love and don't love what you shouldn't love. There's a separate thing from Aquinas (ordo caritatis) which is about how to prioritize limited amounts of care and charity. But in general the Vance style tradcaths are reinventing Gallicanism from first principles.
> 31: Related: this is all fun to think about, but very early polling for the 2028 Democratic primary suggests that by far the #1 candidate is . . . Kamala Harris at 37%, beating Mayor Pete, Gavin, and AOC with 11%, 9%, and 7% respectively. I know you’re not supposed to take early polls like this seriously in terms of who will actually win, but can you take them seriously as a guide to whether people have learned any lessons / no longer love losing? Maybe this is all just name recognition? Also, significant chance that Harris runs for (and wins) the California governorship in 2026.
Yes? This is surprising? I think it's a bad move but I don't see why it's surprising. No one thinks she got a fair shake, she has deep institutional power from the Biden administration, she's got Biden's people to wrangle the groups, and she checks a lot of diversity boxes. Also tons of name recognition. Also the Democrats are really quite bad at doing candidate quality because they're not very good at modeling people outside their bubble. (Republicans aren't great at it either but that's less because they don't understand and more because of purity spiraling around Trump.) It might not last but that's about what I'd expect.
There's a faction of the Democratic party that's going to rally behind her no matter what, and a much larger faction that will go wherever the first faction leads. If that first faction isn't neutralized somehow, I don't think the party as a whole will change course.
I think it's more existential than that. The Democrats effectively need to go through a purge to change ideological and political course. Not a Stalin style shoot the old guard kind of purge but a systematic removal and replacement of large numbers of staffers etc. I think a lot of the current staff would rather stay in place and risk losing.
The most ironic possibility is that Trump saves the Democrats by rooting out all the DEI people. I don't think this will actually work but it *could*.
It depends on the Democrats and how long it takes for there to be moderates storming things. But yes, there's absolutely a world where the Democratic Party reforms quickly and beats JD Vance in 2028. There's also a world where they don't a ride a .5% advantage to beat him. And a world where they're in the wilderness for three or four cycles like they were last time this happened.
This isn't really hard analysis but my feeling is that the rightward shift is real and the Democrats will adapt to it. Whether sooner or later I can't say.
I agree, and "purge" is a word I've used. But I don't think it's precisely existential: they could maintain their previous course until the Republicans again run a candidate bad enough to lose. (They don't have to be good, they just have to be better than the alternative.)
I'm not clear on how much of this is driven by staffers, vs. driven by activists who would call the staffers bad words if the staffers kicked them out.
I think there's no strong division between the two. And I think it's existential for those staffers, not the party.
C.f. Tony Blair's New Labour.
https://xcancel.com/cremieuxrecueil/status/1892095298517766280#m
People don't tend to nominate losers. Al Gore wasn't the nominee in 2024. Kerry wasn't the nominee in 2008. McCain wasn't the nominee in 2012. Etc.
It's highly unlikely that Harris gets nominated again. I'd be surprised if she even runs.
Yeah, "it might not last but it's what I'd expect" at the end gestures towards broadly agreeing with this.
"Tend" is doing interesting work here. Trump lost in 2020, and got re-nominated anyway. (I mean, yeah, a lot of people probably still think he "lost", so maybe that one doesn't count?)
Before 2020, "tend" is working quite well. The last time a candidate lost the general election (major candidate; I won't count third parties) and got nominated again anyway was Nixon in 1968. And it worked! And before that was Adlai Stevenson, who ran in 1952 and again in 1956. Before that, we have to go all the way back to William Jennings Bryan in 1900, who got re-nominated twice, and lost three times. (He even lost the nomination once before regaining it!) Arguably, this is what set the trend, although I recall only two re-attempts in the previous century (Jackson in 1828; Clay in 1844).
More of a mixed bag for VPs that lost. Dole tried in 1996. Mondale in 1984. OTOH, FDR in 1932.
So losing is a fairly big negative, but still surpassable.
> 44: My list of links to publish today includes something like a dozen about DeepSeek, which now seems so thoroughly yesterday’s news that I’m tempted to throw them all out. But in case you still have questions about it, I felt most enlightened by takes from Dean Ball (X), Helen Toner (X), and Miles Brundage (X). The story seems to be that DeepSeek was just very smart, did a great job scrounging up chips from before the export controls hit + mediocre chips that got through the export controls, and did an amazing job wringing as much performance from them as possible. Also, OpenAI delayed announcing o1 for a long time (remember the rumors about “Q*” and “Strawberry”?) and DeepSeek was very fast to announce r1, which made DeepSeek seem less far behind OpenAI than they really were (although this is a comparatively minor consideration - they genuinely did a great job). The absolute worst response to this (from an arms race point of view) would be to give up on export controls - if a rival has geniuses who can use resources ultra-effectively, you don’t want to also give them more resources!
I believe China is purposefully trying to emphasize how the export controls didn't work because they are working and it wants to convince people to lift them. The Chinese government is objectively coordinating with DeepSeek after the launch. The claim they didn't at all before is often repeated but has little evidence.
wonder if you've read this https://stratechery.com/2025/ai-promise-and-chip-precariousness/
author ends with arguing that the US should lift the chip ban, though it's worth noting implicit in his arguments are: (1) AI takeoff is not going to happen soon; (2) Taiwan remaining independent from China in the long-run is unlikely.
we can disagree with his assumptions, but insofar as we think of the future probabilistically, it's worth taking understanding seriously his arguments
Yes, I've heard these arguments. I don't agree with them.
Firstly, a purely realist US still has a strong interest in the Americas but America's immediate environment. That includes Pacific islands like Taiwan.
Secondly, the idea that the chip industry is acting as a shield that will get China not to invade implies that the Chinese government has preserving their tech industry as a goal higher than national reunification. There is no evidence for this proposition and it's mostly advanced by people who are projecting their own beliefs. In fact China has stomped on its own tech industry for nationalist goals several times.
Thirdly, you need to assume that China's going to succeed for a lot of this logic to make sense. China might. But that's by no means guaranteed. Maybe their military won't be all conquering. Or maybe they won't keep up technologically. The closest parallels from the Cold War showed that the Warsaw Pact didn't. They were stuck with increasingly antiquiated technology and, after a decade or two of trying, eventually gave up and began to supply their needs through stealing and copying even more. Which never fully worked out for them.
Lastly, and perhaps most importantly, China was trying to indigenize chip production even before the ban. If you let chips back in they wouldn't stop that effort. The idea that China's attempt to develop this domestically was encouraged by the ban is inventing a causation that does not exist. So the option is whether we want China trying to make its own chips while buying ours (and they will ban ours as soon as they have something comparable) or if we want them doing it without ours.
It’s incredibly unlikely that China starts a war over Taiwan, China is very cautious about using military force, we saw during COVID that they are *too* cautious about safety risks, and will probably be increasingly so due to demographics, and even if it did the ability to develop chatbots and such doesn’t seem to solve for the difficulties of such an invasion. An invasion seems much more likely if China has a weakened economy or mass unemployment due to being outcompeted by US AI and wants to roll the dice on a military adventure to get the chips or at least stop the US from having them too.
So we should ignore what China says (which is that they will pursue reunification at any cost, including using force) and assume we know better than the Chinese know themselves. And in fact the way to make them less aggressive is to continue making them more powerful by making sure they don't fall too far behind in economic competition. And of course, China has no reciprocal obligation to assist the US.
This is an unserious argument and only appeals to people who already either dislike the US or think the US should benefit China.
The export controls are an attempt to monopolize the technology of the future by what is already the richest country in the world, which will create a future of even greater inequality between nations and more desperate people trying to get into the US and leaving themselves at the mercy of Trump.
The entire world should work with China, DeepSeek, and other people creating open-source AI outside those export controls to make sure the export controls fail and the benefits of AI can be shared broadly.
This seems like you're naive about China and blinded by hatred. Whether of Trump or the US I can't say.
And those are bad things because...?
> 49: Deforestation in the Amazon has halved in the last few years (and is down ~75% from its peak). Note that this is only a slower rate of change - total forest coverage is still declining
Pictured: the transition from Jair "I Love Steak" Bolsanaro to Lula "Save The Trees" Da Silva.
This also has to be nominative determinism, right? You can't just have a guy surnamed 'From The Jungle' running on a platform of reducing jungle clearing.
I mean, it's "of the Forest" isn't it? But yeah, Bolsonaro being (I think) battering ram is also kind of funny.
I figured that forest was ‘floresta’ or ‘bosque’ and ‘Silva’ was jungle. I know it derives from the Latin for forest, but in Spanish as well, ‘selva’ means jungle and ‘bosque’ means forest, even as ‘silvestre’ means from the forest.
Apparently it means "bramble, bush"
https://en.wiktionary.org/wiki/silva#Portuguese
One of the annoying things with Spanish/Portuguese/Italian is how often you run into false friends that are so close in meaning they still make sense in context.
It's really weird how people outside Brazil call him Lula da Silva. Not a knock on you, all foreign media does this.
Literally no one in Brazil calls him that. It's just Lula.
If Lula is *not* someone you frequently discuss in the way that cements a name to mean a very specific person, just calling him "Lula" is slightly confusing, since it can also be a woman's first name. "Lula da Silva" makes it clear we're talking about a surname, and this specific person.
I don't think I ever met a woman called Lula. Plenty of Lulu's though.
Adding "da Silva" to his name doesn't really help with disambiguation. If you're talking about heads of state and say "Lula" there's only one.
If you don't know who that is, adding the surname doesn't help.
Names ending in -a tend to be feminine in English. Tolkien uses this as to why Frodo Baggins is called that in English, where his Westron name was Maura Labingi ("Maura" to an English-speaker reads and sounds like a female name, and indeed it *is* an female name in English).
It's the same in Portuguese! "Lula" means squid and is indeed a feminine word.
It's the same as Spanish, where basically every word has a gender.
Words ending in -o are usually masculine, while ending in -a usually feminine.
Lula's a nickname though, so if he was called that from birth it would be weird for a man.
Yeah, I'm aware. Also I needed a surname for the "nickname in the middle" gag to work.
That's a pretty good reason to add it!
To be honest I was worried if the "I love steak" (a nod to Bolsonaro's defense of cattle farmers who are doing a fair bit of the deforesting) was too subtle. But I'll live my weirdness.
Isn't that just recognition level? For a long time in the UK, "Boris" was universally recognised to mean Boris Johnson (except for people with a friend called Boris, I assume), but I'd have been surprised if the same were true in Brazil.
> 50: Lots of buzz over Aella’s appearance on the Whatever Podcast. I haven’t seen it because I don’t watch podcasts, but relevant excerpt here (X), full episode here. I was most interested in Maxwell Foley’s description (X) of the Whatever Podcast’s premise: A "Christian paleoconservative" "debates" OnlyFans models for EIGHT HOURS on whether or not it's bad to have OnlyFans / be a slut, & the women sit through it because they know men watching it will subscribe to their OnlyFans after. Does this qualify as “markets in everything”?
Yes. It's a symbiotic relationship where the hot girls draw people into the podcast and the podcast gives the hot girls a largely young male audience who're they're primary target. If you wanted to do a less anti-woman version of it you absolutely could, by the way. The fact it grew out of the manosphere stuff was largely by coincidence. Lots of the hosts used to do gold digger pranks and stuff and kept with that general vibe. It also makes it easier on the girls to produce viral/blow up content by catering to the audience and means the girls don't need any special skills besides getting yelled at.
The only load bearing parts is it needs to be appealing to men and involve attractive women. So while it can't be scoldy or uptight it could absolutely be something less hostile. There's plenty of ways to have women in bikinis talking about male interest topics that aren't about how women are awful.
> The fact it grew out of the manosphere stuff was largely by coincidence.
Is it? I would expect that men in this sphere are more sexually desperate, and also much more tolerant of blatant attempts to sell using sex.
Is it blatant? It seems like a show about how women suck isn't the most blatant way to sell sex. And I agree it needs to be a large group of sexually desperate men. But I don't think the manosphere has a monopoly on them.
If everyone can see the quid pro quo that's going on, yeah, that is pretty blatant.
Can everyone? I can but I'm not sure how obvious it is.
"There's plenty of ways to have women in bikinis talking about male interest topics that aren't about how women are awful."
Exactly, such as the Chick In a Hot Tub Plays Video Game subgenre.
Yeah. You can't get around the overt sexiness, the appeal to men's interests, the fantasy of it, etc. But I'm not sure you'd want to. Men should be allowed to have interests and their desires and there's nothing wrong with that. And it's certainly healthier than (to quote one conservative commentator) the "This Stupid Bitch" view of women that these manosphere podcasts do.
Yeah, agreed. I do worry sometimes about the rise of parasocial relationships, but that's not limited either to sexy people or the internet (though no doubt correlated with both). I'm liberal enough to say that if a guy wants to pay to watch a girl in a bikini play video games, and a girl wants to make money by doing that, fair enough.
On #5 (good/evil good/bad axis generalizes in AI) -- if this intrigues you, I would recommend meditating on Iris Murdoch's "The Sovereignty of Good over Other Concepts" [https://en.wikipedia.org/wiki/The_Sovereignty_of_Good]. Maybe there's something to the Platonic idea that good things all in some way participate in The Good.
22. Obviously the cost saving is negligible, but you could send everyone a few thousand dollars and they'll be happy. The real benefit IS the decimation of state capacity and the purging of high-IQ civil servants. Causing "chaos and misery for government employees" is the point!
As Russell Vought is supposed to have said, "We want the bureaucrats to be traumatically affected. … because they are increasingly viewed as the villains."
I think no matter how libertarian you are, you still would like the people in the existing government positions to be competent rather than incompetent. Partly this is because some of the existing positions are eg the military, and partly it's because incompetent bureaucrats =/ lazy bureaucrats and they may be more likely to do stupid things than bureaucrats who are smart enough to know why they're bad.
I think your model of the civil service is too nonpartisan. Sure, if they're neutral, you might prefer them competent, but if they're enemies, you want them too incompetent to effectively act against your interests.
I think there are both components. The civil service will occasionally act in a partisan way, but the average civil service task is something like "determining the logistics of drivers license renewals". Yes, there are ways to partisanize this (should illegal immigrants get drivers licenses), but these are a small fringe around the overwhelming bit where people are just better off if they're competent.
(and the average sexy/important civil service task is stuff like running the military, NASA, giving grants to medical research, etc - where again, there are ways to make it more or less partisan, but the core function is something where you want competence)
I lean Democrat but I would rather civil service posts go to Republicans with IQ 110 than Democrats with IQ 90, if those were the only options.
Would libertarians actually want this? Assuming our hypothetical libertarians believe that most government positions don't add much value in the first place, allocating larger numbers of the 110+ IQ proportion of the population to government jobs takes them away from the more value-adding positions in the private sector. I think Musk has expressed similar sentiments, namely that fired government employees can find more productive things to do
This may be the difference between the minarchist and anarchist factions: the former would want a few government jobs done well, and the latter would want them done badly if he couldn't get rid of them altogether.
The former seems a lot more common to me, and the former seems to believe that a great many currently existing government jobs aren't necessary
You don't want IRS people, or spy organizations, or gain of function researches to be competent. Competency and motivation only increases the damage they can inflict. Government employees tend far more in this direction on average.
You definitely want gain of function researchers to be competent! Making "better" viruses is way easier than keeping them in the lab once you've made them.
An incompetent gain of function researcher will make crappy viruses by just trying to get other animals sick with SARS and calling it a day. A good gain of function researchers will be willing to try novel and effective ways of creating all sorts of horrible new viruses. They may be better at keeping it from escaping the lab, but no one is perfect.
When Elon fired a bunch of nuclear safety inspectors, what partisan action was he fearing from them?
They were rehired soon afterwards, so presumably, that wasn't intentional and they just got caught up in a larger set of firings.
One wonders what would have happened if they had decided not to come back. Don’t you think firing people and then re-hiring them again a day later is a little like a mock execution?
There were like 300 of them so only a few would have had to come back to hold the fort (probably while being paid ginormous amounts for overtime) while replacements could be hired and trained
>Don’t you think firing people and then re-hiring them again a day later is a little like a mock execution?
Not at all. I think the reaction of most people who get fired and then rehired because "oops our bad" would be "oh thank god I don't have to look for another job"
Then you'd have to implement whatever contingency plan you have for if they decide to resign en masse.
That's a bit dramatic, but I take your point: I agree this fire-and-rehire thing is not a pleasant experience to subject people to pointlessly. I don't advocate doing this as a matter of policy, and especially not to people you intend to retain.
Fine, what partisan action was he fearing from the air traffic controllers? Does the government being good at making airplanes not crash into each other benefit the left somehow?
What partisan action was he fearing from the National Park Service? Are park rangers a famously leftist group? Does having worse service at national parks benefit the right somehow?
I can do this all day - Musk has fired people from a lot of government agencies, and most of them it's hard to come up with a way they could directly support the left, even if you really stretch.
Since you mention air traffic controllers, here's a possible threat they could be guarding against: if most government agencies are overwhelmingly staffed with people ideologically hostile to you, they could coordinate a strike against you when it'd most hurt you, so it's sensible to purge them at a time your choosing.
> Does having worse service at national parks benefit the right somehow?
Yes, it lowers taxes.
How competent does a bureaucrat have to be to act against your interests? We’re not talking about highly technical skills here. People fear activists in power even if they are not very competent.
These actions are just politicizing the bureaucracy and ensuring that people will be out for revenge.
You are correct, which is probably why the administration is purging these agencies wholesale instead of just removing the most dangerous employees. And also because this is just easier.
> I think no matter how libertarian you are, you still would like the people in the existing government positions to be competent rather than incompetent.
No, the "spy database" admin should be wildly incompetent
Not sure what you mean, but seems like the bad scenario there is that the spy database has terrible security and leaks to the public internet.
Either failure case is better, if the data degrades, good, if its in the public hands, no cia blackmail and we probably get to hear a few cases of them using it.
"Burn it all down" is a luxury belief. The Republican base is quickly getting a crash course in just how much the government does for them.
> Maybe this is all just name recognition?
In first-past-the-post (choose-one) elections, your vote is wasted if it's not for a top 2 contender. Support and funding will centralize around contenders. If you are sufficiently famous, you start off as a contender and therefore have a fundamental advantage. You probably won't finish any lower than 2nd in an election. Kamala is very famous.
This might have something to do with Donald Trump getting elected. I expect more celebrities to follow in his footsteps.
This is one of many reasons why we need STAR voting (http://starvoting.us/) or [approval w/ jungle primary] or some other non stupid system.
> props to X.AI for their unusual decision to have a non-secret prompt
I don't think they actively published it. I think it was just very easy to prompt hack it into spilling the details (and also showed internal monologues about how it wasn't allowed to consider Elon)
I’ve seen Xai engineers on X claim that it’s deliberately easy to prompt hack and they don’t go to effort to hide it.
Re 10: "Last month I linked Sam Harris’ claim that Elon Musk seemed to change into a different person around the start of the pandemic."
Musk has said he uses ketamine on a bi-weekly basis to control his depression (link below).
https://www.cnn.com/2024/03/18/tech/elon-musk-ketamine-use-don-lemon-interview/index.html
And somewhere, I read also said he was taking Anadrol (but I can't find a link for that).
Long-term ketamine abuse is associated with cognitive impairment, including memory loss, decreased attention span, and difficulty with problem-solving.
And ketamine abuse has been linked to mood disorders such as depression, anxiety, and psychosis.
If he's taking steroids that could radically change his behavior.
I doubt if the LSD, mushrooms, and cannabis that he said he frequently consumes would affect his behavior in such a radical way. But my bet is it's either ketamine or steroids.
My impression is that the low doses of ketamine used in normal psychiatric practice (I prescribe it sometimes) don't cause the sort of cognitive impairment we're seeing. Also, I'm not convince Musk is cognitively impaired (in terms of eg ability to solve math problems) so much as manic.
I think it's plausible either that he's taking too high a dose and it's interfering with sleep or something, or that he was latent bipolar and the ketamine pushes him towards hypomania.
Steroids?
Surprised nobody is hypothesizing something like speed (dexedrine, adderall) to stay up and then Ambien to come down. His public actions have a lot of manic/grandiose + I've been awake too long to think straight. And his tweets frequently seem to come from Ambien Walrus Land.
Anabolic steroid use is IMO (as a former long-term user and knowing many other users) extremely unlikely to explain Musk's behaviour. Typically one would experience a mixture of slightly increased drive, feelings of well-being and somewhat increased temper/risk-taking from the typically-used hormones. I've never encountered significant and long-lasting personality changes, not least of the kind that would - all else equal - make one renege on a debt and spectacularly turn against a former friend in the way Harris describes.
Besides, even if he is using steroids, odds are it'd be a (possibly very generous) TRT dose, which has minimal side effects and risk of mental changes, not the likes of Anadrol which are usually used by bodybuilders/athletes, not middle-aged tech bros.
Stopped caring what the pope thought a long time ago. Hopefully he will be replaced by a less toxic, less communist, less creepy version soon.
If you're not Catholic, why did you ever care what the Pope thought? And if you are Catholic, isn't it some kind of sin to gleefully look forward to the Pope's death?
I'd imagine a number of Catholics back in the day were supportive of the Holy Roman Emperor's practice of deposing popes that crossed lines he did not care for. The distinction between them and later Protestants was rejecting the office of the Pope entirely versus just putting a new one in the old one's place.
And the popes often thought the supporters of the Emperor were committing a deep sin.
Doesn't mean much if you get kicked out of the Holy See!
Not if you believe that the Vatican has been taken over by heretics. Honestly, it's only a matter of time before American Catholics build their own Holy See here in the states.
https://en.wikipedia.org/wiki/Sedevacantism to the rescue!
I'm not even joking by the way. With Vance at the helm, I'm confident they can make it happen. Especially if the Vatican elects a black pope. Now THAT would be fun to watch.
The next Avignon Papacy (Baltimore Papacy?) is going to be interesting for sure
To the extent that organized religion is supposed to counter "progressive" excesses, having a pope who's competent at his job is in the interest of even non-Catholic conservatives.
I was a catholic. I don't care about what the church calls sin, especially with regards to the opinions of a very flawed man. I don't gleefully look towards his death. I just assume, perhaps mistakenly, that the world will be a better place after it occurs. He's a bad man.
I found the comparison to Exodus to be amusing, considering God's command when the immigrating Israelites arrived in their new country.
What, specifically, would make the pope a *communist*, instead of just leftlib on some issues and conservative (on the scale of general Western public opinion, not on the scale of Catholicism) on others?
If you think he is toxic and communist, wait until you hear about the Jesus guy who inspired him.
For 13:
If dropping the odds 10% is being done solely to account for the chance of AI unexpectedly overturning the board, then although it might be the right thing to do in some sense, it probably loses to Nate Silver.
Like suppose that there's a 10% chance that a strong AI comes into being and causes all these predictions to be false (or maybe there's a 20% chance that an AI randomizes all of the outcomes). Accounting for this properly should increase your *expected* log-odds score, but since the improvements are correlated, it will usually make your log-odds score worse.
In particular, there's a 10% chance of the AI coming and then you do *much* better than Nate as you assigned every prediction a 10% lower score and none of them came to pass. However, the other 90% of the time, there's no AI and Nate's predictions were (presumably) well calibrated conditioned on there being no unexpected AI takeover. This means that the fact that you biased all of your predictions towards false means you probably did somewhat worse than him. Not enough to balance out the huge improvement to log-odds score that you get if the AI does happen, but enough that you lose handily most of the time.
I don't mean bias all of them to false, I mean bias all of them to 50%, ie Nate's 90% and 10% become my 80% and 20%.
OK. So like a 20% chance that AI randomizes the outcomes (though really this should mean that a probability p becomes 0.8*p + 0.1, so like 60% becomes 58%).
But even this way my logic still applies. If you are right about the chances of AI and its effects (and if Nate wasn't already taking them into account), then you should have a higher *expected* log odds score, but the way that this plays out is that if AI happens, your log odds are much much better, but in the likely event of no AI, your log odds score is worse.
So if your goal is to beat Nate Silver, this probably doesn't do it.
Thanks for the shout out for Codebuff! https://www.codebuff.com/
We'll be doing our 1.0 launch next month. I think we're close to having the best coding agent on the market for existing codebases.
> Gene Smith
> CRISPR
Astonished you didn't point out the priceless nominative determinism.
+1
Was about to say the same thing.
It's a deliberately funny pseudonym he uses for his genetics writing.
Ah. Alas. Thanks!
27: that was a good article (on a great Substack, check it out) at Cartoons Hate Her - glad it’s getting exposure.
One thing, though: my gut feeling is that people sleep with their kids’ nanny way more than all you skeptics think. I am close with two (!!) couples that happened to. One of my wife’s oldest friend’s husband cheated on her with the nanny. Five years later and they’re still together (the ex-husband and the nanny, I mean. My wife’s friend divorced his ass). And an old friend of mine’s wife cheated on him with their nanny. She left my friend and married the nanny - that was over 10 years ago and they’re still together.
Two times among a pretty small circle is a LOT of cuckolding going on with the help. It’s not like I know tons of people with nannies to begin with.
Interesting! I can't think of anyone in my parents' generation or mine who has ever slept with their nanny/babysitter (except I guess a poly person who hired one of their existing partners to babysit, but that doesn't count) and I know a lot of people with nannies/babysitters. What social class are you in?
Shouldn't you append *that you know of* to the end of that sentence? Presumably the the set of people who sleep with their nanny and get caught is a subset of the people who sleep with their nanny, and I don't know what heuristic we should use to guess at the overall size of that group.
I’m an attorney, and we do OK - solid Long Island. But the two couples I’m describing are all over the map.
Lifetime infidelity "base rates" are huge in the few studies that exist. Think ~50% for both genders in modern times. I've never seen a breakdown by class, but even if it's 5x less prevalent for the UMC / PMC, that's still a pretty decent chance.
In-this-relationship base rates are ~20% / 25% for women / men.
For the most solid studies methodologically like NHSLS and NATSAL, the in-this-relationship numbers go down to roughly 10-15% / 15-25% - women / men. but we know those are biased downwards, because only 5% of people reported this when interviewed with somebody else in the room vs 17% if interviewed alone, and the majority were interviewed with somebody else in the room.
Agreed, I linked to this article (https://evoke.ie/2025/02/22/entertainment/stars-accused-affair-nanny) and also found this one (https://www.sheknows.com/entertainment/slideshow/3611/celebrity-nanny-affairs/).
Admittedly a lot of these men are merely wealthy as opposed to "upper-class" (in the sense of intergenerationally wealthy), but the point still stands.
#21 Has anyone explored the angle that the reason the Pope is grandstanding about accepting illegal immigrants (whilst the Vatican continues to have strict policies) is to increase the number of Catholics in the US?
The incoming population seems to be roughly 30-40% Catholic, compared to the US's 20%, so it would be an increase in the number of US Catholics and an increase in their share of the total population.
It's a reasonable suspicion but I don't think it's correct.
Seems you're looking at it from the perspective of "why would the Pope want more open immigration?" You can instead look at it from the perspective of "why would the Pope support immigration restrictions," he doesn't have a clear reason to do so. The justifications immigration restrictionists give for restricting immigration are not generally the kinds of things the Catholic ideology is particularly enthusiastic about. (eugenics, racial purity, Western culture, economic growth, high intelligence, political stability, etc.)
I don't think Popes are that strategic. Also, the Pope seems to be mildly pro (or at least not against) accepting Muslim refugees in Europe, which is against his strategic interests.
If Muslim immigrants officially convert to Christianity then they usually get offered asylum in Europe. Which is obviously a terrible system that is easily gamed.
https://christianconcern.com/comment/fake-conversions-and-asylum-seekers/
You wouldn't believe how many Iranian converts to Christianity Canada gets every year. It's basically a 100% surefire way to get permanent residence in Canada with relatively little effort
How genuine is it?
Obviously bureaucrats (or anyone really) can't definitively prove that someone is faking matters of the heart on a case by case basis, but I could, say, note the fact that despite seeing countless such cases, the Iranian Christian churches in Canada are not overflowing with attendees, even the ones that regularly write letters for alleged Christian converts. I could also note that many Iranians just happen to convert when their visas run out or their visa application is rejected
Information is very easily shared these days among diaspora communities. The Iranian convert story is a very well-known "trick" at this point
I cannot speak for Canada here, but Germany, where this also happens with refugees from Muslim countries. I usually think it is genuine. These are people who are often traumatized, have not had the best experience with the Muslim institutions/communities of their home countries and now arrive to a new country. Often they are welcomed with open arms by Christian communities and get interested in the faith.
In fact, I think it is a very immigrant thing to join a new faith/confession, it seems to have happened all the time in the past.
It isn't genuine, people respond to incentives.
That is a good counterpoint.
A Catholic immigrating from Venezuela to the US means the US has one more Catholic but Venezuela has one fewer Catholic. The net number of Catholics is unchanged, and presumably the Pope is concerned with increasing the number of Catholics, not redistributing them. Venezuela has a higher birthrate than the US does...
The US is the single largest funder of the Catholic church, sending ~13 million USD per year. With a Catholic population of 60 million, this is 0.22/year per capita.
Venezuela's Catholic population is 20 million. Yet the country did not crack the list of top ten donors to the Vatican. The lowest on that list was Canada with ~400,000 USD per year, so Venezuela donates at most 0.02/year per capita, and possibly much lower.
Moving donors from countries with less wealth to countries with more wealth will increase donations.
https://international.la-croix.com/news/religion/vatican-reveals-countries-that-give-the-most-to-peters-pence/16255
On the other hand, if you assume Venezuela's going to remain a Catholic country under any plausible levels of emigration, you might think it good to get more Venezuelans to move to the USA, which isn't Catholic but could conceivably become so if enough Catholics immigrate.
Religious gerrymandering!
Why would any Pope be against illegal immigration? The Pope has no loyalty to the United States. In fact, by Vance’s own definition of Ordo Amoris the Pope should put the interests of his Catholic flock ahead of the interests of non Catholic Americans.
> one could justify a few really good programs like PEPFAR on their own terms (ie if it didn’t exist, I would be outraged until it did)
I was nodding along until I stumbled upon this sentence which... seems far-fetched? Are there programs about the non-existence of which you were outraged as of January 19th 2025? As in, I'm convinced PEPFAR is net-good, but how many people would discuss it in a counter-factual 2025 where Bush never started it?
Given that I'm currently a big fan of efforts to eradicate malaria, I don't it's far-fetched to think that in the alternate world where nobody was working on giving easily-available AIDS drugs to Africans, and AIDS was killing more people in Africa than malaria is now (which I think is likely in this counterfactual), I would be pushing for someone to give AIDS drugs to Africans.
Claude estimates that without PEPFAR there would be around 300,000 deaths under age 24 every year. Currently 6-6.5 million people under age 24 die every year and that number might be as high as 7 million in the absence of PEPFAR, or 7-8% higher.
It also estimates that peak spending on PEPFAR was around $10b/year, adjusting for inflation. When asked if we could save another 300k people under age 24 for $10B it says yes and gives this summary:
> $10B annual charitable funding could prevent ~300,000 under-24 deaths through:
> - $3B neonatal/maternal health
> - $2.5B infectious disease prevention
> - $1.5B nutrition
> - $2B healthcare systems
> - $1B injury prevention
>
> Proven interventions cost $1,000-25,000 per life saved with track records similar to PEPFAR.
But… there’s a lot less talk about this than about PEPFAR and no one is tracking lives lost as a result. This is why I’m ~80% confident we would not be outraged by PEPFAR never existing compared to the funding being cut off today.